The Official 20170614 builds are out

[Josa]

Previous Nightly Builds thread 20170613

The Official 20170614 NIGHTLY builds are out
The Official Firefox 55 Beta 1 builds are out

Previous NIGHTLY: 20170613 (Fx 56.0a1)
Previous BETA: 20170609 (Fx 54.0 RC3)

Tinderbox Builds
Treeherder
Release Notes: NIGHTLY | BETA
Changelogs: NIGHTLY | BETA


Fixed:

1. NIGHTLY +85
2. #1364866 [Core:Audio/Video: Playback]-MP3 parser should honour MAX_SKIPPED_BYTES during findFirstFrame() as well [All][]
3. #1368489 [Core:Audio/Video: Playback]-The event order of onexit and onenter with the same timestamp. [Uns][]
4. #1371127 [Core:Audio/Video: Playback]-Use GetCurrentVirtualThread for MediaManager mMainThreadCheck [Uns][]
5. #1371488 [Core:CSS Parsing and Computation]-stylo: @import rule which fails to load leads to failing of constructing @import rule wrapper object [Uns][]
6. #1371955 [Core:CSS Parsing and Computation]-stylo: Figure out why we reflow the whole document when restyling the HTML spec when we shouldn't. [Uns][]
7. #1368240 [Core:CSS Parsing and Computation]-stylo: Try to do better than eRestyle_Subtree for changes to selectors with descendant combinators. [Uns][]
8. #1370793 [Core:CSS Parsing and Computation]-stylo: mochitest crash - browser_CTP_favorfallback.js [Lin][]
9. #1371125 [Core:DOM: Content Processes]-Don't save JSContext in nsSameProcessAsyncMessageBase [Uns][]
10. #1365402 [Core:DOM: Core & HTML]-textarea value property does not normalize CRLF and CR to LF [Uns][]
11. #1369761 [Core:DOM: Security]-ContextualIdentityService should not be initialized before first paint [Uns][[userContextId][domsecurity-backlog]]
12. #1365097 [Core:DOM]-Convert NS_GetCurrentThread uses in dom (except for dom/media) [Uns][]
13. #1252211 [Core:DOM]-Convert remaining XUL DOM classes to WebIDL [All][btpp-active]
14. #1372371 [Core:DOM]-Fix enumerability handling in the window's resolve hook [Uns][]
15. #1370705 [Core:DOM]-Move side effects of setting attribute of HTMLImageElement to after the attribute is set [Uns][]
16. #1371116 [Core:DOM]-Remove nsJSEnvironment::sContext [Uns][]
17. #1367458 [Core:DOM]-TabGroup::IsBackground() takes 300ms in the content process with telemetry enabled when running [Uns][[qf:p1]]
18. #1370329 [Core:Gecko Profiler]-Allow using PROFILER_LABEL macros from outside libxul, and do so from patched_LdrLoadDll() [Uns][]
19. #1369097 [Core:Graphics: Layers]-After terminating a GPU process youtube videos remain in an infinite loading state [Win][[gfx-noted]]
20. #1372321 [Core:Graphics: WebRender]-Make the clip rect and mask rect relative to the content rect when pushing a clip item [Uns][[gfx-noted]]
21. #1370891 [Core:Graphics: WebRender]-Render blob images in the thread pool [Uns][]
22. #1370879 [Core:Graphics: WebRender]-Share the thread pool between webrender instances. [Uns][]
23. #1362245 [Core:Graphics: WebRender]-Stop running commands during recording for BlobImage [Uns][[gfx-noted]]
24. #1372225 [Core:Graphics]-Add recording for IntoLumaSource. [All][]
25. #1365099 [Core:Graphics]-Convert NS_GetCurrentThread uses in gfx/ [Uns][[gfx-noted]]
26. #1371527 [Core:Graphics]-Create a LayersMessageUtils from GfxMessageUtils [Uns][[gfx-noted]]
27. #1371434 [Core:Graphics]-GLX: Use the proper values when asking for a core context [Uns][]
28. #1372232 [Core:Graphics]-GetPathForGlyphs: Implement default case. [All][]
29. #1088760 [Core:Graphics]-Remove nsRenderingContext [All][]
30. #1369552 [Core:Graphics]-Rename DrawTargetRecording to DrawTargetWrapAndRecord [Uns][]
31. #1371132 [Core:HTML: Parser]-Convert NS_GetCurrentThread calls in parser/ [Uns][]
32. #1371129 [Core:IPC]-Convert NS_GetCurrentThread in IPCStreamSource.cpp [Uns][]
33. #1371117 [Core:IPC]-Remove a use of NS_GetCurrentThread in PBackground [Uns][]
34. #1371079 [Core:IPC]-Use GetCurrentVirtualThread for MessageChannel thread safety checks [Uns][]
35. #1261841 [Core:Internationalization]-(encoding_rs) Replace uconv with encoding_rs [Uns][]
36. #1371216 [Core:JavaScript Engine]-Wasm: Make thread running the ModuleGenerator also perform compilation [All][]
37. #1359874 [Core:MFBT]-Span to Rust slice conversion shouldn't be a footgun for zero-length spans [Uns][]
38. #1365101 [Core:Networking]-Convert NS_GetCurrentThread uses in netwerk/ [Uns][[necko-would-take]]
39. #1361703 [Core:Security: Process Sandboxing]-Enable libevent (i.e. epoll) in linux sandbox [Uns][sblc3]
40. #1370737 [Core:Serializers]-Crash in [@ nsPlainTextSerializer::AppendElementStart] [Lin][]
41. #1362858 [Core:Spelling checker]-Spell-checker goes into self-rescheduling loop and keeps main thread busy spellchecking minified JS for sites like Facebook, Google Docs/Google Sheets due to apparent mozInlineSpellWordUtil::BuildSoftText backtracking. [Uns][]
42. #1371841 [Core:WebRTC: Signaling]-PeerConnectionImpl::RecordEndOfCallTelemetry sends telemetry when no connection information was exchanged [Uns][]
43. #1341285 [Core:WebRTC]-Update WebRTC code to webrtc.org stable branch 57 [All][]
44. #1372063 [Core:Widget: Win32]-Crash in OOM | large | mozalloc_abort | mozalloc_handle_oom | moz_xrealloc | nsTArray_base<T>::EnsureCapacity<T> | nsTArray_Impl<T>::AppendElements<T> | nsDataObj::CStream::OnDataAvailable [Win][]
45. #1369694 [Core:Widget: Win32]-[TSF] TSFTextStore shouldn't cache pref until first use [Win][[qf:p1][tpi:+]]
46. #1361164 [Core:XPCOM]-Add Get{Current,Main}ThreadEventTarget functions [Uns][]
47. #1365098 [Core:XPCOM]-Convert NS_GetCurrentThread uses in dom/media [Uns][]
48. #1365102 [Core:XPCOM]-Convert NS_GetCurrentThread uses in xpcom/ [Uns][]
49. #1370644 [Core:XPCOM]-Deadlock detector seems to pay no attention to the fact that a lock gets unlocked [Uns][[qf:p1]]
50. #1372422 [Core:XPCOM]-mozilla::HangMonitor::IsUIMessageWaiting should always return false in content [Win][]
51. #1372295 [Core:XPConnect]-Call SetLocationForGlobal in a single location in the loader [Uns][]
52. #1371119 [Core:XPConnect]-Load JS prefs per context rather than per runtime [Uns][]
53. #1372382 [Firefox:Activity Streams: General]-Add prefs panel and update strings in Activity Stream system add-on [Uns][]
54. #1370268 [Firefox:Developer Tools: Console]-Stop using Immutable.Record for messages types [Uns][[console-html]]
55. #987877 [Firefox:Developer Tools: Inspector]-[markup view] Add "Copy XPath" context menu item for nodes [All][]
56. #1360957 [Firefox:Developer Tools: JSON Viewer]-Loading failed for the <script> with source “resource://devtools/client/shared/vendor/react-dev.js” [Uns][]
57. #1357964 [Firefox:Developer Tools: Netmonitor]-Fix network-details-panel-toggle direction in vertical mode [Uns][]
58. #1367424 [Firefox:Developer Tools]-[devtools-addon] Use DevToolsShim in addon-sdk [Uns][]
59. #1367425 [Firefox:Developer Tools]-[devtools-addon] Use DevToolsShim in web extensions [Uns][]
60. #861716 [Firefox:Device Permissions]-Multiple async calls to getUserMedia before user response causes earlier requests to be lost [All][[getUserMedia][blocking-gum-]]
61. #1360282 [Firefox:General]-Remove the panel that opens when clicking on the sidebar toolbar button [Uns][[photon-structure]]
62. #1372326 [Firefox:General]-record services initialized during startup [Uns][]
63. #1369640 [Firefox:Preferences]-The search icon in the search input field for Windows and Linux should match the spec [All][[photon-preference]]
64. #1371860 [Firefox:Search]-The 'Search' placeholder should appear immediately on about:home and about:newtab [Uns][[photon-performance]]
65. #692295 [Firefox:Shell Integration]-nsWindowsShellService's ShortcutMaintenance can be removed ~ 6/12 [Win][]
66. #1368383 [Firefox:Sync]-"send to device" subview on page action menu should trigger clients engine sync showing an animation, and update after the sync [Uns][]
67. #1370985 [Firefox:Sync]-Change getBatched() to return records directly instead of via a callback [Uns][]
68. #1368145 [Firefox:Sync]-Send to device context and page action menu should say Send to All Devices [Uns][]
69. #1371962 [Firefox:Tabbed Browser]-Misplaced tab separators near hovered/focused tabs whose titles are faded out to the opposite of the browser's UI direction [Uns][]
70. #1371834 [Firefox:Theme]-Add new Photon-themed download icon [Uns][[photon-animation]]
71. #1372005 [Firefox:Theme]-Update about:rights to new in-content style and add page title [All][[photon-visual]]
72. #1367242 [Firefox:Theme]-[Photon] Implement new styling for header and search box of bookmarks/history sidebar. [Uns][[photon-visual][p1]]
73. #1354186 [Toolkit:Add-ons Manager]-AddonManager.jsm's typeListeners addonListeners and managerListeners arrays should be Sets [Uns][triaged]
74. #1357300 [Toolkit:Add-ons Manager]-Flash plugin preferences are not accessible when it is set to "never activate" [Uns][[good-first-bug], triaged]
75. #1370576 [Toolkit:Application Update]-[e10s] Unusable nightly after updating with a staged update [Win][]
76. #1365895 [Toolkit:Form Manager]-When a select element is autofilled, it should fire events properly [Uns][[form autofill:M3] ETA:612]
77. #1371136 [Toolkit:Performance Monitoring]-Don't use global context in performance monitoring service [Uns][]
78. #1371710 [Toolkit:Places]-Places should not be initialized before first paint [Uns][]
79. #1372152 [Toolkit:Startup and Profile System]-GetShellFolderPath should use SHGetKnownFolderPath instead of SHGetKnownFolderIDList [Uns][[photon-performance]]
80. #1335343 [Toolkit:Telemetry]-Default histogram recording to main & content process [Uns][[measurement:client]]
81. #1358431 [Toolkit:WebExtensions: Frontend]-Add-on name bolded with delay in permissions pop-up [Win][[permissions][triaged]]
82. #1354109 [Toolkit:WebExtensions: Frontend]-Update webextensions code to put things in the permanent area in the overflow panel instead of the hamburger menu [Uns][[photon-structure], triaged]
83. #1287852 [Toolkit:WebExtensions: General]-Mark tabs.highlight as unsupported in the schema [Uns][triaged]
84. #1368545 [Toolkit:WebExtensions: General]-Prevent the ExtensionPreferencesManager from mistakenly overriding a user set preference [Uns][[chrome_settings_overrides] triaged]
85. #1369782 [Toolkit:WebExtensions: General]-Require user interaction for downloads.open() [Uns][triaged]
86. #1370263 [Toolkit:WebExtensions: Request Handling]-Crash in mozilla::extensions::URLInfo::Scheme [All][[clouseau][triaged]]

1. Partial Landings/Diagnostic Patches:
2. #1294025 [Core:Plug-ins]-Intermittent browser/base/content/test/plugins/browser_CTP_crashreporting.js | Uncaught exception - Timed out waiting for plugin binding to be in success state - timed out after 50 tries. [Uns][[stockwell needswork][test disabled]]
3. #1368600 [Core:Security: Process Sandboxing]-Firefox isn't launched with e10s+sandbox after updating it [Win][sb+]
4. #1370890 [Core:Security: PSM]-Upgrade Firefox 56 to NSS 3.32 [Uns][[psm-assigned]]

1. Merge -> Beta 1 Changelog +5
2. #1370087 [Core:DOM]-Crash in mozilla::dom::StorageUtils::GenerateOriginKey [Win][]
3. #1368600 [Core:Security: Process Sandboxing]-Firefox isn't launched with e10s+sandbox after updating it [Win][sb+]
4. #1371962 [Firefox:Tabbed Browser]-Misplaced tab separators near hovered/focused tabs whose titles are faded out to the opposite of the browser's UI direction [Uns][]
5. #1371995 [Firefox:Tabbed Browser]-"New Tab" is cut off / doesn't fade out in loading tabs [All][]
6. #1370263 [Toolkit:WebExtensions: Request Handling]-Crash in mozilla::extensions::URLInfo::Scheme [All][[clouseau][triaged]]

Nightly 56 fixes since 20170613 (Gecko 55) ~106
Beta 55 fixes since 20170306 (Gecko 54) ~4961

Nightly Blog
Release Calendar
Release Tracking
MozRegression Guide

[Omega X]

They launched Fx54 with multiple processes. But dom.ipc.processcount still = 1. Unless I missed the memo somewhere, this is probably a bug.

[PadaV4]

https://wiki.mozilla.org/E10s/Status/June2

The Release Plan is to activate e10s multi for 80% of our Release users without add-ons

[trustjmh]

They launched Fx54 with multiple processes. But dom.ipc.processcount still = 1. Unless I missed the memo somewhere, this is probably a bug.

dom.ipc.processCount.web is automatic setting. With dom.ipc.processcount for user custom setting.

[joeg]

It says above, "The Official Firefox 55 Beta 1 builds are not yet out".

However, in this folder "https://archive.mozilla.org/pub/firefox ... andidates/", there are already five builds of 55 beta.

I know that candidates are not releases (we've been through the discussion about "official" ad nauseam), but I'm still a bit surprised that there have already been five builds.

Anyway, of greatest personal relevance, three of my most important extensions don't work.

I think my beta-testing days for Fx are past tense. Hello ESR.

But hang on ...

https://www.mozilla.org/en-US/firefox/o ... tions/faq/

Does this mean that ESR 52.3.0 is gonna have the changes that are now in Fx55? I see now that I really don't understand the infographic.

Thank you.

[Mouse5]

afaik , the candidates folder/dir is just a test build folder/dir , not an actual release folder/dir .

52 is the Branch of 52, whats in 55 wont be in the 52 ESR Branch

[joeg]

... 52 is the Branch of 52, whats in 55 wont be in the 52 ESR Branch

Thank you for your reply.

Assuming what you're saying is correct, I still don't understand the graphic. I mean, what is the purpose of the vertical downward-pointing arrows? To me it looks like they somehow "link" Fx52 to ESR 52.0 and Fx59 to ESR 52.7.0, the date of the latter being sometime in Q1 2018.

And do you know in which version of ESR the changes that are now in Fx 55 will appear?

Thank you.

[James]

Firefox 55.0 Beta 1 is scheduled for June 14.

Assuming what you're saying is correct, I still don't understand the graphic. I mean, what is the purpose of the vertical downward-pointing arrows? To me it looks like they somehow "link" Fx52 to ESR 52.0 and Fx59 to ESR 52.7.0, the date of the latter being sometime in Q1 2018.

And do you know in which version of ESR the changes that are now in Fx 55 will appear?

The ESR channel is for security fixes and stability as they do not plan to put any new features in ESR that was put in a Release channel build.

Firefox 52.0 ESR is basically the 52.0 Release but with security and allowed stability fixes since. It generally gets eight updates so 52.8.0esr will be out when Firefox 60.0 is Released.

The graph is just saying that 59.0 ESR will be out when 52.7.0esr will be as there is a over lap just like as there has been ever since.

Maybe this calendar will help you understand. https://wiki.mozilla.org/RapidRelease/Calendar

You can see the changes that the 52.2.0esr released today has at https://www.mozilla.org/firefox/52.2.0/releasenotes/

https://www.mozilla.org/firefox/releases/

[PadaV4]

Security patches get patched through to the ESR version. Everything else doesn't. ESR 52.3.0 will get any security patches in FF55 and nothing more.
Edit: James beat me to it..

[joeg]

My thanks to James and PadaV4 - and again to Arch5.

So does this mean that ESR will never get the new UI and changes regarding webextensions?

[James]

So does this mean that ESR will never get the new UI and changes regarding webextensions?

I added more stuff.

Those are features and since adding features is not really a stable thing to do it will not be until Firefox 59 ESR that one would see new features if you use only ESR channel builds. Not to say that some Preferences or tweaks to fix things in the case of say security, crash prone or stability will never happen.

The ESR channel (starting at version 10.0 ESR) was originally made for business and schools and such in mind that wanted stability and not needing to be frequently testing each new Major Release. More regular users may be aware of ESR now days because of reasons like Windows XP and Vista being dropped as of Firefox 53.0 and later versions for example as those users were switched to the Firefox 52 ESR channel.

[joeg]

Thanks. And when is ESR 59 expected?

[James]

As the calendar I linked to https://wiki.mozilla.org/RapidRelease/Calendar says it will be March 6, 2018.

[Omega X]

They launched Fx54 with multiple processes. But dom.ipc.processcount still = 1. Unless I missed the memo somewhere, this is probably a bug.

dom.ipc.processCount.web is automatic setting. With dom.ipc.processcount for user custom setting.

Currently locked by a script that only enables multi-e10s if you have no addons installed. Excellent.

People are going to see "multiple processes, WE are fast now" and the moment they install an addon, it won't be.

[Lurtz]

ESR is not really a recommendable channel for the average user. It gets only security fixes and is more likely to be less stable than the normal release channel.

They launched Fx54 with multiple processes. But dom.ipc.processcount still = 1. Unless I missed the memo somewhere, this is probably a bug.

dom.ipc.processCount.web is automatic setting. With dom.ipc.processcount for user custom setting.

What does dom.ipc.processCount.web do anyway?

It's also pretty tough how Peter Kasting destroys Mozilla's reasoning in the comments there:
https://medium.com/mozilla-tech/the-sea ... 520506aa35