Notices / Schedule
- everything's signed off, releasing today
- proposed schedule at links above, tree is open, please work on blockers
- may adjust schedule to go for early January release to update users with fix for bug 534090 which affects major updates
Firefox 3.6 Beta Testers
- approaching 600,000 ADUs
- getting good feedback through various channels
- Beta 5 builds are being tested by QA now, targeting a Thursday release unless we get to RC first
Firefox 3.6 Release Candidate
- we are really, really close to being code-complete (see blocker report)
- only need 8 more patches, and a tracemonkey merge
- need all hands on deck for those patches and on incoming nominations
- if we can go to build today or tomorrow, QA will scrap Beta 5 and we'll release RC to the beta audience ASAP
Firefox 3.6 RC Code Completion
- 4 nominations to be triaged
- 25 code blockers remaining:
- Release Blockers (flag: blocking1.9.2 or blocking-firefox3.6)
Browser / Front End
(Progress reports every weekend on Planet Firefox)
- see our active projects and get involved / propose others
- mozilla-1.9.3 front end development (see planet firefox for quick overview):
- Drew's made Places data visible in a content window
- Blair's tweaking tab match in awesomebar, should land on trunk soon
- Rob Strong is fixing the updater for unprivileged windows users, and finding wins in JS startup time (see more)
- 2 blockers
- As usual, subpixel positioning bug is the long pole, but we won't block 3.6 on it. No ETA on that bug.
- 1 nom
- 2010 Q1 GFX goals - Performance!
- Looking to land direct2d acceleration on mozilla-central "before Christmas."
- We have a Talos test for scrolling that Jeff's working on; ETA also before Christmas. bug 534819
- 1.9.2 report
- No layout blockers, no layout noms
- 1 widget blocker (roc)
- Reviewing syncbase timing for SMIL (birtles)
- Making new assertions fatal on Tinderbox? (dbaron)
- Exhortation to tackle orange bugs using record+replay
- One 1.9.2 blocker left, fixed on trunk.
- Refactored to represent each JSObject with an actor pair.
- frameLoader.attachChildGlobal(node, "propertyName") to create a property on |node| corresponding to the child's global object, thereby exposing the wrapper.
- Grass-roots support (mrkbap, peterv) for a change in terminology: Cross-Process Object Wrappers (CPOWs)?
- Clone-documents-for-printing landed (smaug)
- HTML5 parser work moving along, talos problem figured out (problem in talos page set, not in the parser).
Platform-specific Support Update
- Most blockers fixed on TM. Newly added debug builds are orange, will fix today.
- continuing to land performance improvements
- Read the latest weekly update. (none last week)
- View snapshot results for startup and all other tests.
|Estimated win||Bug #||Summary||Owner||Status||Notes|
|>10%||bug 512584||Super fast paths for Components.classes and Components.interfaces||Taras Glek||in progress||Got consultation from mrbkap, needs new patch.|
|>10%||bug 525013||Investigate a more static build configuration of Firefox||Joel Reymont||Mac and Linux in progress||Only measured on Mac so far, need Linux numbers and someone to figure out the Windows story.|
|up to 25%||bug 514083||Per-file HFS+ compression on Mac OSX 10.6||Joel Reymont||in progress||Snow Leopard only.|
|TBD||bug 520309||Startup cache: replacement for fastload cache||Ben Hsieh||review||More efficient than current fastload, and key to enabling fastloading of XBL, CSS, manifests and various other data. Need to push to Places branch to figure out base perf difference from current fastload.|
|TBD||bug 503483||Turn on --enable-faststart for Firefox by default||Dietrich||needs testing||Loads Firefox core libraries at boot time. Need to test on all OSes, publish the numbers, and get discussion going.|
|TBD||bug 513149||Speed up CSS parsing||Zack Weinberg||Zack's blocked on other work||Taras says it's high, but no figures, so need better data here.|
- OOPP has landed and appears to have stuck!
- Minor issues abound (IPC on only): e.g. any plugins installed to a path with spaces are broken
- Need to get Talos changes landed (bug 529137) so that we can enable OOPP by default on E10s tree, and then once blockers are fixed, on mozilla-central (blocker tracking).
To turn on IPC plugins, set the pref dom.ipc.plugins.enabled in nightlies starting this morning.
- Stopped sending test results to *-Unittest trees. bug 534252 Scraping will need to be turned on as new columns show up.
- Enabled mochitest-ipcplugins tests on m-c, places, tracemonkey bug 530522
- Turned on optimized and debug packaged tests on tracemonkey bug 524176
- Still need to do extended downtime so IT can do database maintenance bug 528573. All trees that report to talos will be need to be closed.
- Need to discuss adding a null deref to make add-ons fail early, i.e., Relevant Knowledge uses the a11y API in a non-threadsafe way. We should block the dlls, but also keep add-on developers from using this in a way that will eventually crash. The problem is we don't know how many add-ons this will break (and will it matter?). Also, is there a patch ready to make this happen? If not, can we get one ready so that we are prepared to pull the trigger? Peterv has the details. (damons)