Platform/2009-04-07: Difference between revisions

From MozillaWiki
Jump to navigation Jump to search
Line 6: Line 6:


'''[[Releases/Firefox 3.0.10|Firefox 3.0.10]]'''
'''[[Releases/Firefox 3.0.10|Firefox 3.0.10]]'''
* Scheduling a release for mid-late May
* Please work on your [https://bugzilla.mozilla.org/buglist.cgi?keywords_type=nowords&keywords=fixed1.9.0.10+verified1.9.0.10&field0-0-0=flagtypes.name&type0-0-0=equals&value0-0-0=blocking1.9.0.10%2B&field1-0-0=flagtypes.name&type1-0-0=notsubstring&value1-0-0=approval1.9.0.10&order=map_assigned_to.login_name,bugs.bug_id blockers]
* Went through nomination list yesterday
* Code freeze is in two weeks
* Blocker owners will be emailed today


'''[[Releases/Firefox_3.5b4|Firefox 3.5 Beta 4]]'''
'''[[Releases/Firefox_3.5b4|Firefox 3.5 Beta 4]]'''

Revision as of 17:59, 7 April 2009

« previous week | index | next week »

Notices / Schedule

Firefox 3.0.9

  • Aiming for April 21 release

Firefox 3.0.10

  • Please work on your blockers
  • Code freeze is in two weeks

Firefox 3.5 Beta 4

  • string frozen
    • any additional strings?
    • New Tab?
    • Geolocation?
  • code freeze: Wednesday, April 15th
  • QA start: Thursday, April 16th
  • release: late in week of April 20th

Blocker Report

Since 1.9.1 branch ... 2009-04-07-blocker-report.png

Past 2 weeks ... 2009-04-07-blocker-report-detail.png

[ Platform Blocker Queries | Front End Blocker Queries ]

Summary

  • blockers going down again (yay!)
  • moving more fixes over to branch (yay!)
  • we're staying on top of blocker triage
  • approval triage fell off a little

The Breakdown

Browser / Front End

  • Polish update: Firefox is 51% shiny (+2% over last week)
    • Thanks go to: rstrong, jim (polish hard), and Dao (4)
    • 46 remaining easy polish bugs (whiteboard [polish-easy])
    • 36 remaining hard polish bugs(whiteboard [polish-hard])

pub?key=pMZGKUlD9NOPg4oJGs1CUMw&oid=1&output=image&foo=.png

GFX 1.9.1 Update

  • Bug status:
    • 5 blocking 1.9.1+
      • 3 landing soon
      • 1 gfxTextRun crash, being tracked down by jdaggett and roc
      • bug 470487 - the nsWindow plugin destroying crash.

Layout 1.9.1 Update

  • 1.9.1 query: x unfixed 1.9.1 blockers
  • Layout
    • x blockers (6 with patches or probably will be fixed by other bugs)
    • No beta4 blockers
    • 4 untriaged noms
  • SVG
    • x blockers
    • No beta4 blockers
  • Video/Audio
    • x blockers (most have patches or work-in-progress)
    • Media cache blocks beta4, but is ready to land
    • Video perf improvements (720p running smoothly, probably just one bug away from smooth 1080p on reasonably decent machines)
  • Beyond 1.9.1
    • Plugin testing patches done, moving on to actual compositor hacking!

Content Update

  • 1.9.1:
    • 12 blockers
      • bug 483915 Workers: Use new JSAPI to limit tracing code cache size -- Trivial, but needs bug 474497 first.
      • bug 484107 Security bug. -- Has patch, needs small tweak.
      • bug 480713 nsDocLoader doesn't clear requests from mRequestInfoHash,... -- Ready to land.
      • bug 481566 Content sink needs to be more responsive -- Did a touch-and-go, but needs updated patch.
      • bug 483818 Leak when script appended via DOM does document.write -- Needs patch
      • bug 462853 "Assertion failure: slot < (obj)->map->freeslot" during T... -- Needs more work, hard to reproduce, might not block on this.
      • bug 486269 Security bug -- needs review.
      • bug 485125 Crash on close [@ libflashplayer.so@0x4110d ] -- Likely fixed, or fixed enough, by Karl's recent gtk widget destruction fix.
      • bug 475864 Quickstubs allow pages to mess with native anonymous content
      • bug 479560 Security bug (mrbkaps wrapper work) -- Making good progress
      • bug 480205 Security bug (mrbkaps wrapper work) -- Making good progress
      • bug 483672 Permission denied for <http://localhost:7080> to call met... -- Needs reviews, somewhat scary. Firebug workaround also a possibility here.

JS 1.9.1

  • 31 blockers
    • x P1s not! including upvar bug 452498. Tryserver builds of upvar passed linux/mac perf, mochi, reftest, crashtest. Failed TUnit, browser-chrome, with what looks like a single bug.
    • x P2s
  • 1 untriaged nomination

General 1.9.1

These are bugs that fall outside of components covered by the Gfx, Content, Layout and JS groups:

Mobile 1.9.1 Update

  • Big perf things we need to land soon
    • bug 485426 ARM simd rgb24->565 optimizations
    • bug 482676 Neon pixman optimizations
    • bug 483409 Use XShm for pushing to the screen
    • bug 482689 Optimize background displaylists for opaque images
    • bug 67752 Interruptible Reflow
      • We would really like to see this land on trunk as soon as possible so we can try to decide if it could go back to 1.9.1 for mobile.
  • In progress

Security

Security Reviews

Still outstanding / to be scheduled

  • native JSON
  • DNS prefetching

STOP HERE AND CHECK AGAINST SCHEDULE

Tree Management

  • Crash stacks for tests:
  • Greener-Tinderbox: work ongoing, machines still running. Updating to new changeset this week. Will have more results this week - many less reftest failures now that the color depth is set properly on the win32 boxes.

Roundtable