The Official Win32 20100212 Trunk build won't be out.

[Peter(6)]

Nightly TRUNK thread RSS feed (courtesy DaCypher)

.The Official Win32 20100212 Trunk build won't be out.

Previous Trunk nightly 20100211

Hourly Trunk builds (Minefield next ...): mozilla-central-linux || mozilla-central-macosx || mozilla-central-win32 || Firefox Tinderbox || Trunk Changelog
Hourly archive for Windows, Mac and Linux (not the final format yet)

Last update: 2010-02-12 -- 23:59 PST = 07:59 UTC
The tree is CLOSED
Last Changeset: 93c7b23284b8

Fixed:

1. #468810 [Firefox:Search]-Make "Manage search engines" window resizable [Win]
2. #523083 [Core:DOM: Core & HTML]-Traverse/unlink cached content list [All]
3. #531225 [Core:DOM]-Workers: Share strings across thread boundary [All]
4. #531801 [Core:Networking: HTTP]-Should cache SSL content to disk even without Cache-Control: public [All]
5. #534527 [Core:Disability Access APIs]-Accessibility needs new DOM API [All]
6. #538918 [Core:Plug-ins]-[OOPP] Modal system dialogs cause UI freeze w/windowless controls [Win]
7. #540692 [addons.mozilla.org:Blocklisting]-Firefox Crash [@ vksaver.dll@0x3d09 ] and various other addresses [Win]
8. #541924 [Core:Graphics]-bad underline font list doesn't work [Win]
9. #542620 [Core:Layout: Tables]-table sometimes ignore background-color of column [Win]
10. #543788 [Core:Plug-ins]-[OOPP] CallUpdateWindow hang when switching tabs with a plugin loaded [Win]
11. #543901 [Core:Plug-ins]-plugin process crash with null NPPluginFuncs::getvalue [@ @0x0 | mozilla::plugins::PluginInstanceChild::AnswerNPP_GetValue_NPPVpluginNeedsXEmbed(bool*, NPError*) ] [Lin]
12. #543961 [Core:Disability Access APIs]-nsAccessibilityService::GetAccessible() shouldn't try to get document accessible from global cache twice [All]
13. #544882 [Core:IPC]-youtube videos won't play unless mouse is actively moving [Lin]
14. #544999 [Firefox:Theme]-New toolbarbutton style for the main window [Win]
15. #545001 [Core:Geolocation]-NetworkGeolocationProvider should support HELD protocol [All]
16. #545330 [Core:Print Preview]-Crash [@ PresShell::Destroy] on exit after failed Print Preview of XUL document [Lin]
17. #545368 [Core:Widget: Gtk]-firefox doesn't theme tab focus rings [Lin]
18. #545391 [Core:Disability Access APIs]-keep nsAccessNode objects in cache instead of nsIAccessNode [All]
19. #545502 [Core:Widget: Gtk]-Enable the "Selection" option for GTK's print dialog [Lin]
20. #545506 [Core:Widget: Gtk]-Fix the custom header/footer dialog in GTK's print dialog [Lin]
21. #545550 [Core:SVG]-SVG/SMIL: Setters for SVG attr base value shouldn't directly modify animated value, when animations are active [All]

Partial Landings/WIPs/Incoming:

1. #386937 [Core:GFX: Color Management]-ConvertPixel doesn't work as nsColor and gfx_color are not compatible... [All]
2. #517553 [Core:General]-Remove '=='/'!=' 'PR_FALSE'/'PR_TRUE' [All]
3. #521467 [Core:Networking: HTTP]-Automatically log in to proxy [All]
4. #527659 [Core:Security: PSM]-Update Mozilla-central to NSS 3.12.6 [All]
5. #542731 [Core:SVG]-Code cleanup in /content/smil [All]
6. #545338 [Core:Plug-ins]-RPCChannel should sync NotifyWorkerThread calls with pending messages [Win]
7. #545593 [Core:Layout: View Rendering]-Remove nsViewManager::Enable/DisableRefresh. [All]
   
   
8. #526411 [Core:Video/Audio]-after pausing a video it often won't restart playing later [Lin]

Regressions/Annoying/Common bugs:

1. #423128 [Firefox:Places]-favicons aren't backed up to *.json [All]
2. #539658 [Core:Plug-ins]-[OOPP] flash doesn't work in full-screen mode [Win]

Trunk fixes since 20090813 (mozilla 1.9.3) = ~ 2366
Preliminary roadmap: Firefox 4.0 beta 1 Q3,2010 - release Q4,2010 (source)
Mozilla Developer Preview 3.7 alpha1 [20100210 2294]

[JCRUYFF]

this version 3.7 when will become to public will re-numbered to 4.0? because mozilla say dropped 3.7


PD: Sorry for my Bad English

[RedFlames]

this version 3.7 when will become to public will re-numbered to 4.0? because mozilla say dropped 3.7


PD: Sorry for my Bad English

It will happen when it happens. I wouldn't worry about it too much as, at the end of the day, it's just a number.

[JCRUYFF]

this version 3.7 when will become to public will re-numbered to 4.0? because mozilla say dropped 3.7


PD: Sorry for my Bad English

It will happen when it happens. I wouldn't worry about it too much as, at the end of the day, it's just a number.

I was hoping that FF 4.0 was come close to Google Chrome 4.0 in Speed



PD: Firefox 3.7a2pre is about 50% slower than Chrome 4.0

[sciguyryan]

It will happen when it happens. I wouldn't worry about it too much as, at the end of the day, it's just a number.

I was hoping that FF 4.0 was come close to Google Chrome 4.0 in Speed



PD: Firefox 3.7a2pre is about 50% slower than Chrome 4.0 [/quote]

1. That is on a benchmark which may or may not be true of the real world test cases.
2. Work is still being done on TraceMonkey to improve it's speed and efficiency.

[teoli2003]

There are quite a lot of work on TraceMonkey (and JaegerMonkey). Some have just landed on the Tracemonkey branch (see Bug 542002 - Optimize to flat closures even if some upvars can't be copied who just landed yesterday!).

I'm convinced that Firefox.next will have faster JS than now (I would expect 30% bettering for Sunspider, so half-way to Opera 10.5, which is the best on my computer, slightly better than V8 in dev build of Chrome or the latest nightly of Nitro in Webkit). 30% is my non-scientific estimation, of course.

[owen.c93]

Can anyone confirm the using ctrl+tab to switch tabs doesn't work with youtube? And possible other flash player sites.

[Izbitzer]

Can anyone confirm the using ctrl+tab to switch tabs doesn't work with youtube? And possible other flash player sites.

That's probably Bug 78414 which is super old. Would be nice to see a fix though.

[Ria]

Ctrl+tab seems to work, but not clicking on other tabs.
It seems even workaround for the hang.

[Andu]

There are quite a lot of work on TraceMonkey (and JaegerMonkey). Some have just landed on the Tracemonkey branch (see Bug 542002 - Optimize to flat closures even if some upvars can't be copied who just landed yesterday!).

I'm convinced that Firefox.next will have faster JS than now (I would expect 30% bettering for Sunspider, so half-way to Opera 10.5, which is the best on my computer, slightly better than V8 in dev build of Chrome or the latest nightly of Nitro in Webkit). 30% is my non-scientific estimation, of course.

Isn't the latest Opera build already using SSE2 optimizations? We really have no clue what happens when FF gets a SSE2 and a D2D/DW implementation.

[exactlyy]

Can anyone confirm the using ctrl+tab to switch tabs doesn't work with youtube? And possible other flash player sites.

That's probably Bug 78414 which is super old. Would be nice to see a fix though.

for me it does work either by using ctrl+number or ctrl+tab.
what doesnt work is trying to click another tab while watching some video on youtube

[teoli2003]

I doubt D2D/DW will have any (major) impact on JS performance (It will on other part of Firefox, but not JS). According SSE2, some of it is already implemented and it will surely help some (especially in x64, where it is sure to have it available), but that surely not it that leads to the difference in Sunspider's results between Chrome, Opera, Safari on one side and Firefox on the other.

[sciguyryan]

Inline JIT is one thing I'm looking forward to seeing, also multi-thread JIT which I believe is also on the table. Those should help. Does anyone know if recursion is traced in TraceMonkey yet?

[Andu]

Shouldn't there be a JIT/TM meta bug somewhere?

[owen.c93]

Can anyone confirm the using ctrl+tab to switch tabs doesn't work with youtube? And possible other flash player sites.

That's probably Bug 78414 which is super old. Would be nice to see a fix though.

Only 9 years old! Thanks.