My Browser Builds (Part 3)

[roytam1]

New regular/weekly KM-Goanna release:
https://o.rthost.win/kmeleon/KM76.4.5-Goanna-20220122.7z

Changelog:

Out-of-tree changes:
* update Goanna3 to git afea1b3ff...f71f51550:
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1200795, part 1 - Eliminate excessive detail from cycle collector graph memory reporting. r=njn (8775704ad4)
 - Bug 1200795, part 2 - Add memory reporting for CCGraph::mPtrToNodeMap. r=njn (154912392e)
 - Bug 1202670 - Move most of nsCycleCollector::RemoveObjectFromGraph into CCGraph. r=smaug (78b449602d)
 - Bug 1163006, part 1 - Make the cycle collector use the concrete logger class. r=smaug (a23f4adfd3)
 - Bug 1163657 - Clean up setting of mFlags in CCGraphBuilder's ctor. r=smaug (c49fa66d20)
 - Bug 1163006, part 2 - Rename various cycle collector listener variables to logger. r=smaug (c655503d1b)
 - Bug 1163006, part 3 - Clean up some cycle collector logger set up code. r=smaug (9825ec8e80)
 - Bug 1163006, part 4 - De-COM the nsICycleCollectorListener methods we only call from C++. r=smaug (39e579782f)
 - Bug 1163006, part 5 - Add a less COM-y getter for mWantAllTraces. r=smaug (70c424197e)
 - Bug 1152505 - CCGraphBuilder::NoteChild should pass aEdgeName by reference. r=smaug (25ffabb54c)
 - Bug 1189122 - Assert when we Suspect() when a CC scan is in progress. r=smaug (ae997c0fec)
 - Bug 1189829 - fix quadratic behavior in nsConsoleService; r=erahm (2b37337085)
 - Bug 1189231 - Impl operator->* to nsAutoPtr. r=nfroyd (a491e0e775)
 - Bug 1196050 - Replace NSCAP_Zero usage with decltype(nullptr). r=froydnj (4b4d411103)
 - Bug 1196925 - Remove the non-existing relocatable store buffer counts from about:memory; r=njn (d030d12a81)
 - Bug 1123237 - Part 2. MemoryProfiler hooks in js engine. r=terrence (e3495c3520)
 - Bug 1196847 - Part 1: Allow storage of a unique id for a cell independent of address; r=jonco (53b94846f1)
 - Bug 1207730 - Update comments in Barrier.h and sort classes in same order; r=sfink (ba3abf54cb)
 - Bug 1196847 - Part 2: Implement a cell hasher that uses unique id based hashes; r=jonco (3811c58b7a)
 - Bug 1196848 - Add Zone::uniqueIds to about:memory reports; r=njn (bff35eb98c)
 - Bug 1123237 - Part 3. Monitoring allocation and gc events in nursery and tenured heaps. r=terrence (47cc20bb98)
 - Bug 1123237 - Part 4. Monitoring allocations and frees for ArrayBuffer. r=terrence,sfink (3495e46571)
 - Bug 1123237 - Part 5. Don't emit inline allocation when memory profiler enabled. r=terrence (b4b44652d0)
 - Bug 1123237 - Part 6. A new API to get backtrace without allocating memory in profiler. r=mstange (eff2d42cbd)
 - Bug 1123237 - Part 7. XPCOM interface for memory profiler. r=smaug (f3a85f8181)
 - Bug 1123237 - Part 8. Tracking the memory events. r=BenWa,terrence (8694fa4ad4)
 - Bug 1123237 - Part 9. Interface to memory-profiler add-ons. r=jimb (1f8ed472ad)
 - Bug 1123237 - Part 10. Expose SwapElements from nsBaseHashtable. r=nfroyd (a2e06425ad)
 - Bug 1123237 - Part 11. Don't use STL in memory-profiler. r=BenWa,cervantes (c13096634c)
 - Bug 1123237 - Part 12. Fix GC hazards. r=terrence (64776fc60e)
 - Bug 1194424 - Part 0: Extend the core dump protobuf message format to accomodate allocation stacks; r=sfink (d25ca13aec)
 - Bug 1167292 - Part 1: Add a telemetry probe to time how long it takes to save a heap snapshot; r=ejpbruel (25a9047d18)
 - Bug 1167292 - Part 2: Add a telemetry probe to time how long it takes to read a heap snapshot; r=ejpbruel (71e3bb5fe3)
 - Bug 1167292 - Part 3: Add telemetry probes for node and edge counts when saving heap snapshots; r=ejpbruel (7985b9edf6)
 - Bug 1187062 - Part 0: Make js::Debugger::getObjectAllocationSite return a SavedFrame* rather than a JSObject*; r=sfink (0ec4175b9c)
 - Bug 1204613 - Add a static_assert that ProfileEntry::{Flags,Category} do not overlap; r=djvj (a2d3672676)
 - Bug 1204168 - Add the js::ProfileEntry::setCategory method; r=djvj (dce07b7eb4) (14412dba8)
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1153255 - set a build configuration flag when MOZ_TSAN is enabled; r=jorendorff (6b787252a5)
 - bug 1155829 - remove AccessibleWrap::SetMaiHyperlink r=surkov (69718cc37e)
 - bug 1155829 - remove MaiHyperlink::Initialize r=surkov (cf770ba16e)
 - bug 1155829 - inline MaiHyperlink::GetAtkHyperlink r=surkov (2eb9846a7e)
 - bug 1155829 - add shutdown method to MaiAtkObject r=yzen (f4e35f223d)
 - bug 1155829 - declare MaiAtkObject in nsMai.h r=yzen (fdb5cd60c7)
 - bug 1155829 - move AccessibleWrap::GetMaiHyperlink to MaiAtkObject::GetAtkHyperlink r=surkov (e00ba50a1a)
 - bug 1164976 - move AccessibleWrap::FireStateChangeEvent to be a member of MaiAtkObject r=davidb (5aee8b6119)
 - Bug 1128365 - MOZ_TOOLS is unnecessary. r=glandium (d7c8d8798d)
 - bug 1169376 - Allow getting the OuterDocAccessible for the tab's top level document r=davidb (630780f18e)
 - bug 1171728 - Only look for an OuterDoc accessible parent of a proxy if it doesn't have a proxy parent r=lsocks (76a55151bd)
 - bug 1107337 - Teach refChildCB to deal with proxies r=lsocks, davidb (1a5cfddb63)
 - bug 1172053 - don't call malloc in FireAtkShowHideEvent r=lsocks (350bbd99e1)
 - bug 1172053 - don't malloc the signal name in FireAtkTextChangedEvent r=lsocks (9035d29090)
 - bug 1107337 - support downcasting Accessible to OuterDocAccessible r=lsocks, davidb (d946ce3bf4)
 - bug 1107337 - Add OuterDocAccessible::RemoteChildDoc() r=lsocks, davidb (9fd40de39d)
 - bug 1172523 - Move AccessibleWrap::FireAtkTextChangedEvent to MaiAtkObject::FireTextChangeEvent r=lsocks (021898bef4)
 - bug 1146518 - allow MaiHyperlink to store references to proxies r=surkov (e068e47855)
 - bug 1146518 - make atk hyper link impl support proxies r=surkov (7e4d74efb3)
 - bug 1146518 - create MaiHyperlinks for proxies r=surkov (1b2b0ef5fa)
 - bug 1146518 - Only pass hyper links to MaiHyperlink::MaiHyperlink r=surkov (2c506d17b7)
 - bug 1163070 - fix AtkHyperlinkImpl::getHyperlink after bug 1146518 r=surkov (0a33758e48)
 - bug 1172523 - Allow getting the modified text from an AccTextChangeEvent without copying r=lsocks (aa0bf42335)
 - bug 1170153 - check documents have a docshell before trying to tell the parent process about new remote DocAccessibles r=davidb (8c35f9cec4)
 - bug 1181177 - null check tabChild in DocManager::CreateDocOrRootAccessible r=davidb (6c3e309edb)
 - bug 1186528 - don't create a DocAccessibleChild if we can't tell the parent process about it r=lsocks (d3954843b9)
 - bug 1164976 - fire useful state change and caret move events for proxies r=davidb (ef2e39e6d9)
 - bug 1172525 - rework forwarding events to the parent process r=davidb, lsocks (36537fb8f9) (88653c180)
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1001637 - Make math tables implement the nsIAccessibleTable interface. r=surkov (ebbb4325d6)
 - Bug 1149118 - Expose radio group with the correct role on OS X, r=surkov (4d4a95735c)
 - Bug 1113153 - use 'paragraph' role for paragraphs only, introduce 'text' role for small text containers, r=marcoz (14c4fa5881)
 - Bug 1175164 - Do not expose MathML semantics element in the accessibility tree. r=surkov (3945ae0367)
 - Bug 930843 part.1 Add test for checking isTrust attribute value of DOMActivate event r=smaug (14a49dd447)
 - Bug 930843 part.2 NS_UI_ACTIVATE event should be trusted event even if it's caused by an untrusted event r=smaug (e0af6165b5)
 - Bug 1156315 - Call DispatchInputEvent in more places on OS X. r=dvander (e24f42c09f)
 - Bug 1186015 part.1 Rename nsIMEContext to mozilla::widget::IMEContext r=jimm (077bf6e772)
 - Bug 1186015 part.2 Rename nsIMM32Handler to mozilla::widget::IMMHandler r=jimm+m_kato (a2098e33b5)
 - Bug 1184533 Rewrite range check in aOffset of nsIMM32Handler::GetCharacterRectOfSelectedTextAt() r=emk (fb68bc3c56)
 - Bug 555642 part.1 nsCaret should have a way to override the caret visible state for hiding caret temporarily and nsEditor should hide caret if composition string doesn't have caret information r=roc (9de4aa366c)
 - Bug 1171847 Remove unnecessary ',' in logging code in nsTextStore::Initialize() r=m_kato (78d0d43625)
 - Bug 1181714 nsTextStore should store previous focused document manager until new document manager actually gets focus r=m_kato (fd7e820af2)
 - Bug 1176950 nsTextStore should keep storing locked content until NOTIFY_IME_OF_COMPOSITION_UPDATE is notified r=emk (0a12349575)
 - Bug 1137539 part.1 Implement TSFStaticSink::IsGoogleJapaneseInputActive() r=emk (68d1f9a0ec)
 - Bug 1137539 part.2 Implement TSFStaticSink::IsATOKActive() r=emk (51a4208bfa)
 - Bug 1137539 part.3 Implement TSFStaticSink::IsFreeChangJieActive() r=emk (affbb81b96)
 - Bug 1137539 part.4 Implement TSFStaticSink::IsEasyChangjeiActive() r=emk (abd3e674b6)
 - Bug 1186014 Rename nsTextStore to mozilla::widget::TSFTextStore r=jimm+m_kato (ca63351f94)
 - Bug 1187351 TSFTextStore should forget modified range at notifying TSF of layout change since GetTextExt() shouldn't return TS_E_NOLAYOUT after that r=emk (48878793e0)
 - Bug 1187367 TSFTextStore shouldn't destroy native caret for ATOK until notifying TSF of layout change r=emk (a2a4022d15)
 - Bug 1187566 TSFTextStore::Content should compute mMinTextModified Offset only with the latest composition string and original composition string, and also the hack should be enabled on Win10 r=emk (1368885efe)
 - Bug 1050644 part.1 Add methods to check whether the active TIP is Chinese TIP which deosn't show candidate window in e10s mode r=emk (1b783e682e)
 - Bug 1050644 part.2 Hack ITextStoreACP::GetTextExt()'s offset for some Chinese TIPs of MS r=emk (c05cec10ce)
 - Bug 1187579 Enable TSF in e10s mode r=m_kato (1ae106231a)
 - Bug 555642 part.2 IME handlers on Windows shouldn't append caret range if the caret is in the target clause which doesn't have specific style r=m_kato (88bae2c8da)
 - Bug 1196124 Fix mismatch of printf style string and following arguments of MOZ_LOG() in IMMHandler::GetCharacterRectOfSelectedTextAt() r=m_kato (0f6debd0d3)
 - Bug 90712 Cancel composition when Enter key message isn't consumed by IME r=m_kato (0febdb1fd9)
 - pointer style (6c984274cd)
 - Bug 1177268 - implement aria-rowcount/index and aria-colcount/index, r=marcoz (92ac300374)
 - Bug 1153135 - Return possibility PEN working while APZC is enabled. r=jimm (8bc77ff84f)
 - Bug 1171073 - Suppress content scrolling issue on 64-bit platforms. r=kats (62db0dead3)
 - Bug 1121946, Implement e10 cursor drag feedback on Windows, r=jmathies (6e8218d202)
 - Bug 1144650 - Don't dispatch touch-based mouse events when APZ is handling touch. r=jimm (53ebea0d90)
 - Bug 1163056 - InkCollector activation at first PEN event. r=jimm (b28e0ee58b)
 - Bug 895274 part.3 Make the enum of event messages a named enum IGNORE IDL r=smaug (32e04a4e90)
 - spacing (b9249221dc)
 - Bug 1182966 - Part 3: Add some final annotations to DomainSet. r=mrbkap (3455822e89)
 - Bug 1050122 - Part 1: Enable Nuwa on debug builds, and kill it if the preference is false. r=khuey (46685a7bf3)
 - Bug 1146229: Remove calling convention modifier from local variable. r=surkov (e5f4b6e5d5)
 - bug 1168932 - Implement ProxyCreated and ProxyDestroyed to update mozAccessibles r=tbsaunde (77a007fa21)
 - Bug 1169701 - Fire native OS X accessibility events for proxied accessibles r=tbsaunde (c4847135a8)
 - bug 1172523 - fire useful text change events for proxies r=lsock (85b7f995b8)
 - remove spurious file (229c787360)
 - bug 1171117 - Fix cairo to build on iOS. r=jrmuizel (45709c8fe3)
 - Bug 963738 - Annotate XPCNativeMember, r=terrence (d1afdde52b)
 - Bug 1148383 - Add testcase for previously fixed object metadata issue r=terrence (2c1b762cba)
 - Bug 1149797 - Update browser jstests to treat ecma_6/extensions tests as 1.8, so that let is interpreted as a statement and not as an expression. r=bustage in a CLOSED TREE (ac6127d6bc)
 - Bug 1150858 - Unwrap the correct object in ArrayBuffer.transfer (r=sfink) (cf67906802)
 - Bug 1150380 - IonMonkey: MIPS: Fix JitRuntime::generateProfilerExitFrameTailStub. r=rankov (eeac49ddce)
 - use RootedValue for POW as in 1135708 (34c1c3d191)
 - Bug 1196648: IonMonkey - Don't run the lazy link stub for asmjs to jit fastpath, r=nbp (686d495371)
 - Bug 1121947, Implement e10 cursor drag feedback on Mac, r=smaug (f48df2d1)
 - and sync widget/TextEvents.h and TextRange.h with AF. (b91a6a24c)
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1141906 - Adjust some assertions in Linux sandbox feature detection. r=kang (365e9a6a8e)
 - bug 1169408 - Merge mozButtonAccessible and mozPopupButtonAccessible r=surkov (6996bb22e8)
 - Bug 1162434 - Make windows emulation work on X64, r=marcoz (f0bb96de8a)
 - Bug 1162434 part 2 - Fix ISimpleDOMNode Unique ID on 64 bit systems, r=surkov (1afbaa9fec)
 - code style (3cabaf0882)
 - Bug 734229 - Partially address by refusing to re-negotiate on NTLM. r=mayhemer, r=keeler (59b7feea33)
 - Bug 1164714 - Move netwerk/base/nsISiteSecurityService.idl into security/manager/ssl. r=keeler,mcmanus (154e93a29c)
 - Bug 1164714 - Move netwerk/test/TestSTSParser.cpp into security/manager/ssl/tests/. r=keeler (7fc68a2b66)
 - Bug 1124649 - Part 1 - Add specific error messages for various types of STS and PKP header failures. r=keeler,hurley (e967aef5fc)
 - Bug 1046421 - Do not disclose the system hostname via NTLM handler. r=honzab (a4a85439ca)
 - Bug 1084025 - Add telemetry to measure failures due to not falling back. r=keeler (5388e21c64)
 - Bug 1106470 - Drop SSLv3 support entirely from PSM. r=keeler (5de896acc0)
 - Bug 1195606 - Use channel->ascynOpen2 in security/manager/ssl/nsNSSCallbacks.cpp (r=sicking) (8b5e18f069)
 - Bug 1197644 - Remove the security.ssl.warn_missing_rfc5746 pref. r=keeler (f76ae9af3e)
 - namespace (7817663683) (f71f51550)

* Notice: the changelog above may not always applicable to XULRunner code which K-Meleon uses.

A goanna3 source tree that has kmeleon adaption patch applied is available here: https://github.com/roytam1/palemoon27/tree/kmeleon76

[Mathwiz]

Thanks, @InterLinked! Now I finally understand what a "polyfill" is!

On 1/20/2022 at 4:17 PM, InterLinked said:

Why NM27 and not 28? (though I guess both, the merrier...)

Same reason there are multiple versions of 360EE: the older versions are lighter and thus run better on older hardware.

Which reminds me: your ChromeFill extension would be great for 360EE v11, which is based on Chromium 69. Lots of folks don't want to run the RAM-hogging v13. So you might want to post in the 360EE threads too.

BTW, the globalThis polyfill won't be as useful on NM 27 as on 28 because, of @roytam1's builds, only UXP browsers like NM 28 support Javascript modules. But it's probably still worth doing if possible.

[roytam1]

2 hours ago, Mathwiz said:

globalThis polyfill

do UXP based applications need this? https://github.com/roytam1/UXP/commit/5386cd812b9fe8294a57eae7ed6d2ef117912af4

[UCyborg]

On 1/20/2022 at 1:18 AM, InterLinked said:

Anyways, these id*** had their fun messing up their JavaScript to break all our stuff, but I'm taking my JavaScript back now... take that, Google...

No, you're not.

[UCyborg]

41 minutes ago, roytam1 said:

2 hours ago, Mathwiz said:

globalThis polyfill

do UXP based applications need this? https://github.com/roytam1/UXP/commit/5386cd812b9fe8294a57eae7ed6d2ef117912af4

They added globalThis soon after Reddit stopped working then if I remember correctly.

I don't expect fix for the current breakage to come soon though.

[soggi]

The mirror of latest BNavigator, Firefox 45ESR, IceApe (new!), IceDove (new!), K-Meleon 1.5.x/74/76, MailNews 52, New Moon 26.5/27/28 and Serpent 52/55 builds by @roytam1 has been updated -> soggi.org - tools.

changelog:
- added latest IceApe 52 20220122 build
- added latest IceDove 52 20220122 build
- added latest K-Meleon 76 20220122 build
- added latest New Moon 27 20220122 builds
- added latest New Moon 28 20220122 builds
- added latest Serpent 52 20220122 builds
- added latest Serpent 55 20220122 builds
- as well as find out for yourself

kind regards
soggi

[InterLinked]

5 hours ago, roytam1 said:

do UXP based applications need this? https://github.com/roytam1/UXP/commit/5386cd812b9fe8294a57eae7ed6d2ef117912af4

I don't *think* so but not sure.

New Moon 28 works fine with StackOverflow, which didn't work in Iron 70 properly due to the globalThis issue, so I'd say "no".

This is the compatibility chart for globalThis: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/globalThis#browser_compatibility

There are other areas where UXP is affected, but I haven't addressed those yet.

[InterLinked]

4 hours ago, UCyborg said:

No, you're not.

Which version is this? Seems newer than what I'm using.

I did notice the "Unexpected token ?" error. I took a look and it almost looks like invalid JS to me, they weren't using it as the ternary operator. I'm not sure what the polyfill/fix for this is yet.

Some things I use are still broken, affected by issues other than in or in addition to just globalThis. Hopefully, we can add more support for these other things over time.

[UCyborg]

I don't think you can "polyfill" nullish coalescing operator. It's occurring more frequently so good luck with the modern web on old browsers.

I think Iron from that screen was supposed to be 70. I got this particular installer since it's not in the official archive: https://www.filepuma.com/download/srware_iron_64bit_70.0.3650.0-20779/

Some downloads there are definitely wrong, Iron 70.0.3700 was actually Iron 71.something.

[InterLinked]

5 minutes ago, UCyborg said:

I don't think you can "polyfill" nullish coalescing operator. It's occurring more frequently so good luck with the modern web on old browsers.

I think it could be fixed by processing all the JS and doing some kind of find and replace, maybe.

How to do that, I'm not sure yet.

5 minutes ago, UCyborg said:

I think Iron from that screen was supposed to be 70. I got this particular installer since it's not in the official archive: https://www.filepuma.com/download/srware_iron_64bit_70.0.3650.0-20779/

Some downloads there are definitely wrong, Iron 70.0.3700 was actually Iron 71.something.

Yeah, it *used* to be there but it's not anymore it seems.

It's in the Wayback Archive though - the latest 70 version with the old UI is linked here: https://w2k.phreaknet.org/

[UCyborg]

3 hours ago, InterLinked said:

Yeah, it *used* to be there but it's not anymore it seems.

Actually, the link I posted is actual 70.0.3650.0.

3 hours ago, InterLinked said:

It's in the Wayback Archive though - the latest 70 version with the old UI is linked here: https://w2k.phreaknet.org/

While this one is 70.0.3650.1.

[InterLinked]

6 minutes ago, UCyborg said:

Actually, the link I posted is actual 70.0.3650.0.

While this one is 70.0.3650.1.

 

6 minutes ago, UCyborg said:

Actually, the link I posted is actual 70.0.3650.0.

While this one is 70.0.3650.1.

Yeah... that's the latest 70 version, no? One minor version higher than the other one.

[NotHereToPlayGames]

On 1/20/2022 at 11:55 AM, InterLinked said:

All right, available now!

https://blog.interlinked.us/66/when-the-world-wide-web-goes-on-strike-how-do-you-fight-back

https://github.com/InterLinked1/chromefill

Hands are a bit tied at the moment, but might try to see what will work for New Moon as well.. for now, this unbreaks enough sites on Iron 70 that it's usable enough again.

Bummer, I was hoping this would "unbreak" my water bill web page but it doesn't.

Are these errors polyfill-related?

[InterLinked]

12 minutes ago, NotHereToPlayGames said:

Bummer, I was hoping this would "unbreak" my water bill web page but it doesn't.

Are these errors polyfill-related?

Polyfills fix JS-deficiencies in the browser, it's not likely polyfill related, per se, a polyfill *might* be able to fix it.

I also saw several "XYZ" undefined errors on some things that don't quite work yet. I think a couple more polyfills are needed for this to unbreak close to most sites.

Most helpful errors are in the developer console (F12):

If you go to the Console tab, the raw JavaScript errors are printed out there. Best to use those rather than what's on the page itself, though in this case, it's probably similar by the looks of it. The console will also show you the exact JS that crashes the page.

[Mathwiz]

Looks like an add-on for the globalThis polyfill should be targeted to NM 27, since globalThis is apparently implemented in UXP (NM 28, etc.) already.

Speaking of polyfills, I poked around in JustOff's github-wc-polyfill add-on a bit. It seems to implement custom elements, which has been a JavaScript standard for some time, but still not implemented in UXP.

But his add-on limits the polyfill to GitHub/GitLab. Anyone know why? Since it's a standard, I'd think you'd want to add it to all Web pages, wouldn't you?