Quote
J.G.

Quote
cheetos1897
Cannt load twitter.com
It says 'Something went wrong, but don’t fret — let’s give it another shot'

The error message seems to be something related with failed script:
https://twitter.com/?failedScript=vendors%7Emain
Changing UA does not work for me. :mad:

because they changed javascript that requires gecko 47+ grammar.

Tue Oct 26 2021 09:07:32
Error: SyntaxError: yield expression must be parenthesized
Source file: https://abs.twimg.com/responsive-web/client-web-legacy/vendors~main.9294a885.js
Line: 1, Column: 23361

lets see if this can be workarounded.

Quote
roytam1

Tue Oct 26 2021 09:07:32
Error: SyntaxError: yield expression must be parenthesized
Source file: https://abs.twimg.com/responsive-web/client-web-legacy/vendors~main.9294a885.js
Line: 1, Column: 23361

lets see if this can be workarounded.

Thank you Roytam1.
Does this mean tutanota.com has a chance? I got similar messge when try to login tutanota.com, i.e., parenthesized...

Now the Instagram has stopped working(( . User-agent changed... no

New regular/weekly KM-Goanna release:
https://o.rthost.win/kmeleon/KM76.4.5-Goanna-20211030.7z

Changelog:

Out-of-tree changes:
* update Goanna3 to git 06f122f42...2087e70e1:
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1167248: Test whether mysterious rand_s failures are due to LoadLibraryExW hooks. rs=terrence (4c38bf34b)
- Bug 1176289 - Record proper optimization tracking outcome on failure of getPropTryConstant due to non-singleton value. r=shu (6a718490e)
- Bug 1180013 - Fix irregexp on ARM64. r=efaust (88bb8599a)
- Bug 1163891 - Read lines of input in xpcshell/jsshell while properly handling EINTR failures and retrying in response. r=glandium (c84415384)
- Bug 1176890 - Start running jstests faster on windows; r=sfink (f878247e4)
- No Bug - Fix test count generation when jit-flags is set; r=meow (9f90cbd6e)
- Bug 1177801 - Remove NewObjectWithProto. r=bhackett (64a29265e)
- Bug 1179315 - part 1 - make TypedArrayLayout's constructor inline and MOZ_CONSTEXPR; r=lth (e7da1ad42)
- Bug 1179315 - part 2 - make it more obvious that typeIDs of typed arrays are constants; r=lth (796348a4d)
- Bug 1197316 - Remove PR_snprintf calls in xpcom/. r=froydnj (1c775e4c0)
- Bug 1210607 - Check for null compartment in PopulateReportBlame() r=terrence (4b19494a2)
- Bug 1145636 - Update jsapi-tests/README. (The functional change is where it talks about the Makefile; that was stale.) r=Waldo. (b0cd84037)
- Bug 1123875 - Add a test for strict assignment to a readonly WebIDL attribute. r=peterv. (1c8f5fae8)
- Bug 1145218 - Require a new float manager if a block frame has a different writing-mode to its parent's. r=dbaron (7ebcd8fab)
- Bug 1150151 - Set leak threshold for reftests (r=mccr8) (19cb1a82e)
- Bug 1149781; Update Oculus runtime for new SDK 0.5; r=jrmuizel (aea8b2c31)
- Bug 1149781; [webvr] Update Oculus API to use 0.5.x SDK; Mac startup crash fix; r=dbaron, a=startup-crash (ad874bd1b)
- Bug 1145936 - Fix sizing of foreignObject with vertical writing mode. r=smontagu (059e46e95) (f0504b101)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1118272 - Hold a cpu wakelock for async call "AlarmsManager.add". r=fabrice (2e04b9b47)
- Bug 1146754 - Reftest for selection highlighting of trailing space. r=mats (5b9083192)
- Bug 1146754 - Fix reftests for selected shadowed text to avoid trailing-space issues. r=mats (727f06840)
- Bug 1146754 - Show selection highlighting for trailing space. r=mats (16955b975)
- Bug 1146754 followup - Add a little fuzz to the testcase for OS X 10.10. no_r=me (31efb0691)
- Bug 1148376 - Only force-enable layers acceleration on Gonk. r=jmuizelaar (28bdfa602)
- Bug 1149519 - Workaround for GDI's failure to return empty extents for blank glyphs. r=roc (0467912a9)
- Bug 1174230 - Limit the complexity of regalloc bundles which can be merged together, r=sunfish. (929511dd4)
- Bug 1029483 - Fix various 99ch style violations in Intl.cpp. r=sparky (9f5de7ea9)
- Bug 1093421 - Add a test for support for currencyDisplay: 'name' or 'code' when formatting numbers. r=abargull (1107405c5)
- Bug 946528 - Update the default number of digits of precision in numbers formatted as CLF currency (Chile's Unidad de Fomento) in the Intl API. r=abargull (268fbd219)
- Bug 1177898 - Fix error checking in CloneFunctionObjectIfNotSingleton. (r=efaust) (e8509a49e)
- Bug 866596 - Properly reject locales with duplicate singletons or variants as invalid within the Intl API. r=abargull (edc6917a8)
- Bug 866596 - Properly reject locales with duplicate singletons or variants as invalid within the Intl API. r=abargull (0c0bfe280)
- pointer style (c6953c513)
- Bug 1178976: Add a post-write barrier to ObjectWeakMap::add. r=terrence (89b9a7b2f)
- Bug 1177941 - Do not show warning about unreachable empty statement after return statement. r=Waldo (f661cf694)
- spaces, pointer style (c22647ade)
- Bug 1161351 - Fix unwound exit frame sizes in JitProfilingFrameIterator. (r=nbp) (d5f415602)
- Bug 1172498 - Properly mark ion frame new.target values. (r=jandem) (a2fc125ec)
- Bug 1163520: Make IsInternalFunctionObject take its argument by reference, as it must not be nullptr. r=shu (f02635f28)
- Bug 1163520: Don't hand out internal function objects via Debugger.Environment.prototype.callee. r=shu (13ac48c56) (e39070aba)
- ported changes from mozilla upstream: Bug 1250589 - Remove the must-be-parenthesized requirement from yield expressions. r=waldo (4f327d0de)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1175347 - Add a JSAPI test for exposing availability of locales derived from the default locale. r=itsatest, rs=till over IRL (0bba85db1)
- Bug 1175347 - Fix a test that assumes too much in compacting builds. r=intermittent-orange in a CLOSED TREE (b2992b31c)
- Bug 1171180 - Remove trunc from jsmath; r=jorendorff (0e8913338)
- Bug 1176864 - Truncate modulo operator for int32 r=h4writer, r=nbp (5de9c5340)
- Bug 1166790 - Remove old Statistics formatting code; r=sfink (6abc54dab)
- Bug 1171612 - Use C++11 features to make Statistics module nicer; r=sfink (d0cfef988)
- Bug 1132208 - "Remove dead code from framerate actor now in GraphsWorker.js". r=jsantell (4bfaf2773)
- Bug 1146237 - FramerateActor should use docShell.now() rather than performance.now so that page refreshes do not break it. r=vp (a0d3fbd28)
- Bug 1134079 - Supply audio node definitions directly from the client if the webaudio actor server does not support it (like on older versions of FxOS). r=vp (95f728d47)
- Bug 1172183 - Pull out the implementation of FramerateActor so that it can be consumed by other actors. r=vp (1c3f0d82a)
- Bug 1172182 - Pull out memory utility logic out of the MemoryActor so other actors can consume it. r=vp (851d1cce3)
- Bug 1172182 part 2: correctly link to the memory module via timeline actor. r=vp (3714663db)
- Bug 1172184 - Pull out logic from TimelineActor into a standalone module that can be consumed by other actors. (031127dc0)
- partial of Bug 1159506 - Make GC events use TimeStamp. r=terrence (99fa0378f)
- Bug 1035973 - Add DebuggerObject.getOwnPropertySymbols; r=fitzgen (fd10482d5)
- Bug 1174712 - Tolerate singleton objects with uncacheable prototypes in Ion caches, r=jandem. (7c3ce4fdc)
- add limits for numeric_limits (67b09aa7f)
- Bug 811911 - Use UTF-8 output for TwoByte chars in Error objects; r=jandem (2fd3cf6cd)
- Fix the test for bug 1173787 to work even when the filename contains a ':' (e.g. on Windows) so we can reopen the CLOSED TREE (b42aa3b92)
- Bug 1100498 - Report function names for addon exceptions. r=billm (dfa69d830)
- Bug 1170840 - Add testbed allocator for new regalloc features, and use for a change to hot/cold code bundle splitting, r=sunfish. (95c484a37)
- Bug 1174542 - Remove unnecessary AutoWritabeJitCode from initTraceLogger. r=luke (a427e979c)
- Bug 1173529: IonMonkey - Also iterate phis when removing guards, r=nbp (055a5dcd7)
- Bug 1114079 - Fix overrecursion check in nsGlobalWindow:etNewDocument to not report a JS exception. r=bz (7c0ba0677)
- Bug 1172513 part 1. Fix shell's Evaluate to actually throw when it detects save/load bytecode mismatches. r=waldo (5f5ccc094)
- Bug 1172513 part 2. When XDR-encoding a function, don't incode temporary flags. r=waldo (404fa1939)
- Bug 1171430 - Do not poison the nursery's chunk kind marker; r=terrence (248667fc9)
- Bug 1171430 - Don't shift the poison value by non-byte values; r=jonco (320ba4ba0)
- Bug 1173908 - Fix an MSVC warning about negating an unsigned integer; r=nbp (3244fd56c)
- Bug 1172545 - Recover from OOM if Ion compilation is skipped due to not being able to allocate a 'this' object, r=jandem. (918025215) (00ce1ce6f)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1125481 - Replace FunctionInfo macros by variadic templates. r=Waldo (83244859f)
- Bug 1125481 - FunctionInfo: Remove useless inline keywords. r=Waldo (a13a186af)
- Bug 1169769 - Stop pretending js/src is a top-level directory. r=gps (2ba218bbf)
- Bug 1132189 - Rename "Analysis Helper" thread. r=bhackett (1a6b50af5)
- Bug 1172824: Initialize a few more members in VerifyPreTracer (CID 1304706); r=terrence (9fb0f4b03)
- Bug 1172824: Close dumpFile in case of error in DumpHeap (CID 1302587); r=jonco (e552f4320)
- Bug 1172641 - Fix class tests to not fail on builds where classes are not enabled r=shu (56474b665)
- Bug 1153244 - increase JS helper threads stack space when MOZ_TSAN is enabled; r=bhackett (cb5c9fc01)
- Bug 1172895: Raise the stack size limit for JS helper threads; r=bhackett (750cc5e19)
- Bug 1166101 - redirect aka os.file.redirect should not be available in fuzzing mode (09f574daa)
- Bug 1139376: Ionmonkey - Make removing of guards aware of phis, r=nbp (fb7cb028b)
- Bug 1171722 - Part 1: Add gcreason to GCSliceCallback. r=fitzgen (a9e4ae93b)
- Bug 1171722 - Part 2: Add gcreason:PCONNECT_SHUTDOWN. r=terrence (815cbd630)
- Bug 1171722 - Part 3: Specify gcreason:PCONNECT_SHUTDOWN during xpconnect shutdown. r=mccr8 (5123cda83)
- Bug 1171722 - Part 4: Don't warn if DebuggerOnGCRunnable::Enqueue fails during shutdown. r=mccr8 (efc3b8aab)
- Bug 1173513 - Point to SavedFrame.md docs from JSAPI comments; r=fitzgen (6c9bb84db)
- No Bug - Remove an unneeded annotation DONTBUILD; r=sfink (05fe03101)
- Bug 1171871 - Fix bytecode documentation comment for JSOP_NEWTARGET. r=efaust (bacbef1ce)
- Bug 1169460 - Fix imprecise computation in BacktrackingAllocator::minimalUse, r=sunfish. (7ea535e67)
- Bug 1170959 - Remove JS_GetPropertyDescriptor from a JSAPI test. r=efaust (8745b6d04) (e361c5511)
- ported dependency of "Bug 1174307 - Add some internal content policy types for the purpose of reflecting them on RequestContex", some part may already in the tree:
- Bug 1175114 - Translate the content policy type obtained in CSPService::AsyncOnChannelRedirect to an external one before invoking the content policy implementation; r=smaug
- Bug 1175122 - Add more assertions to the in-tree content policy implementations to ensure that they receive external content policy types; r=baku
- Bug 1178339 - Account for service workers separately in WorkerDomainInfo; r=nsm (4e0540b82)
- ported follow-ups of "Bug 1143922 - Add AsyncOpen2 to nsIChannel and perform security checks when opening a channel - scriptSecurityManager changes (r=sicking,bholley)":
- Bug 1184387 - Bail out of file:// loads for all non-chrome:// URIs. r=Gijs
- Bug 1224046 - Remove <base href> from directory listings. r=mcmanus (dbff70221)
- ported 2 commits from mozilla upstream to fix ask-to-activate plugin issue:
- Bug 1232903 - Skip Security checks if triggeringPrincipal is SystemPrincipal (r=sicking)
- Bug 1257650 - Skip Security checks if triggeringPrincipal is SystemPrincipal only for subresource loads. r=sicking (2087e70e1)

* 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



Edited 1 time(s). Last edit at 10/30/2021 12:49AM by roytam1.

@roytam1 thank you for your work!

Quote
my
Now the Instagram has stopped working(( . User-agent changed... no

Instagram is using new for/in syntax that is not to be parsed correctly by current goanna3 engine.

Problem: Preferences ->
JavaScript ->
Permissions -> You can add a site but you cannot remove a site from the list.


Does not work with versions:
KM76.4.5-Goanna-20211030.7z
KM76.4.5-Goanna-20211023.7z
KM76.4.5-Goanna-20211016.7z
KM76.4.5-Goanna-20211009.7z
KM76.4.5-Goanna-20211002.7z
KM76.4.5-Goanna-20210925.7z


The last properly working version is KM76.4.4-Goanna-20210918

Quote
lilu
Problem: Preferences ->
JavaScript ->
Permissions -> You can add a site but you cannot remove a site from the list.


Does not work with versions:
KM76.4.5-Goanna-20211030.7z
KM76.4.5-Goanna-20211023.7z
KM76.4.5-Goanna-20211016.7z
KM76.4.5-Goanna-20211009.7z
KM76.4.5-Goanna-20211002.7z
KM76.4.5-Goanna-20210925.7z


The last properly working version is KM76.4.4-Goanna-20210918

this will be addressed in next build.

Thank you Roytam1 for your hard working

Crashes when visiting any page of wikipedia.org

Quote
cheetos1897
Crashes when visiting any page of wikipedia.org

can't reproduce here. what build you're using? do you use clean profile?

Crash would not happen if u Only visit the homepage of wikimedia.com. Try some content page.
such as this one
https://en.wikipedia.org/wiki/Olympic_Games

Sometimes KM doesnt crash after the page of wikimedia.com is completly loaded
but if u scoll the page up and down for several times it crashes rightaway.

Tried some recent Builds and it turns out that this issue starts to show since Biuld 20211009

Quote
cheetos1897
Crash would not happen if u Only visit the homepage of wikimedia.com. Try some content page.
such as this one
https://en.wikipedia.org/wiki/Olympic_Games

Sometimes KM doesnt crash after the page of wikimedia.com is completly loaded
but if u scoll the page up and down for several times it crashes rightaway.

Tried some recent Builds and it turns out that this issue starts to show since Biuld 20211009

This does not happen on my PC (build KM76.4.5-Goanna-20211030). Please, if there is someone who has the same problem, report it here so that roytam1 can fix it or, if it is not a build problem, so that cheetos1897 can focus on checking his computer configuration and roytam1 can move forward with the next build.
If we don't collaborate even reporting, then we can't come here to complain that K-Meleon builds are not developed fast enough.
Regards

Quote
cheetos1897
Crash would not happen if u Only visit the homepage of wikimedia.com. Try some content page.
such as this one
https://en.wikipedia.org/wiki/Olympic_Games

Sometimes KM doesnt crash after the page of wikimedia.com is completly loaded
but if u scroll the page up and down for several times it crashes rightaway.

Tried some recent Builds and it turns out that this issue starts to show since Biuld 20211009

Forget to mention that this issue Only happens on Win xp

try this url https://www.okooo.com/soccer/match/1154958/ah/
KM cannot load data properly

Open it with other browsers,like chrome, firefox etc. You will see the difference

Quote
cheetos1897
try this url https://www.okooo.com/soccer/match/1154958/ah/
KM cannot load data properly

Open it with other browsers,like chrome, firefox etc. You will see the difference

website firewall rejected data request and returning HTTP 405, there is nothing I can do.

Having an issue with wwe network. Works on the final xp firefox with the adobe login for canadian users but on kmeleon it for some reason has an endless spinning cycle instead of loading the tv providers

Id use firefox but while it can login to the tv providers properly it cant play any of the videos. Is there a way to get the adobe login to trigger somehow so i can pick my
Provider and login? If not for the canadian tv login it would work as it seems to play the free videos from there. Just can’t login something with the adobe sign in it seems

Hi guys, On the last build Kmeleon dont show IMAGES when I search with Google, at least this version, possible other builds have same problem..... Tested only on Windows 7 Professional. Thanks

Quote
kmbrowser
Hi guys, On the last build Kmeleon dont show IMAGES when I search with Google, at least this version, possible other builds have same problem..... Tested only on Windows 7 Professional. Thanks

It is true. Fortunately this malfunction does not happen on Windows XP, where images are displayed without problem, but it seems that Google has changed something that now affects the display of images using K-Meleon on Goanna on newer versions of Windows (I have tested it with Windows 7 Pro and Windows 10 Pro). This seems to happen with any version of K-Meleon on Goanna on these O.S. where, however, images are displayed using the old K-Meleon on Gecko.

New regular/weekly KM-Goanna release:
https://o.rthost.win/kmeleon/KM76.4.5-Goanna-20211106.7z

Changelog:

In-tree changes:
- browser-omni: update permission API usage after bug 1170200 is landed in goanna3 side (ba00a387)

Out-of-tree changes:
* update Goanna3 to git 2087e70e1...00fee1e0a:
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1170200 - Part 2: Update JavaScript consumers of nsIPermissionManager::Remove; r=ehsan (aa1cbab01) (fd3bc37f6)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Fixup for hazard bustage after bug 1168719 on a CLOSED TREE. r=sfink (9c8d92138)
- Bug 1169869 - Document the Debugger's onPromiseSettled hook DONTBUILD; r=jimb (d601304f1)
- Bug 1168416 - Do not change lastIndex property if not writable in RegExp.prototype.{test,exec}. r=Waldo (3bd43d20b)
- Bug 1171749 - Re-add nonincremental reason to compact stats output; r=sfink (d257c5d3b)
- Bug 1165460 - Work around on-demand linker badness with NDK r10e/GCC 4.9 r=dhylands,glandium (dcdf2c92b)
- Bug 1170673 - MSVC is less flexible with zero-sized arrays; r=shu (e556ce826)
- Bug 1155618 - Add some more OOM tests r=terrence (b6637848b)
- Bug 1155618 - Test OOM behaviour of global creation r=terrence (eac8ae615)
- Bug 1155618 - Don't retry memory allocation if we're simulating OOM r=terrence (990c80307)
- Bug 1155618 - Fix tracelogger destructor that touches possibly uninitialised hash table r=benj (657e587bc)
- Bug 1161303 - Fix bug in NewObjectCache::newObjectFromHit() caused by previous patch in bug 1155618 r=terrence (3132d9d4d)
- Bug 1155618 - Fix some more places where we didn't report OOM to the context r=terrence (f9cf4b194)
- Bug 1155618 - Fix some more places we don't report allocation failure to the context r=terrence (33fa561c9)
- formatting (af5b94792)
- Bug 1054756, part 1 - Support symbol-keyed properties in JSXrayTraits::resolveOwnProperty(). r=jandem. (8f8f8df4b)
- Bug 1054756, part 2 - Move ToPrimitive slow paths into a non-inline function. r=jandem. (6bdde7843)
- Bug 1193924 - Add assertion to AsyncOpen to make sure asyncOpen2() was called first when securityflags in loadInfo are set (r=sicking) (e440e90b6)
- Bug 1194519 - call AsyncOpen2 within child and parent if securityFlags are present (r=jduell) (1ce87ade4)
- Bug 1204648 - Support AsyncOpen2,Open2,nsIUploadChannel and nsIUploadChannel2 on nsSecCheckWrapChannelBase. r=sicking (f6b6ae40f)
- Bug 1204703 - Add AsyncOpen2 to channels within JS tests (r=sicking) (2a650afa0) (45e1308fe)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1051146 - Part 1: Introduce the B2G specific build flags, initially enabled. r=smaug (341c7565d)
- Bug 1051146 - art 2: Introduce the new Pocketsphinx and Sphinxbase code with no build integration. r=smaug, r=gerv (f29eafa78)
- Bug 1051146 - Part 3: Introduce build system changes for the new Pocketsphinx and Sphinxbase code. r=smaug, r=gps (06c60429d)
- Bug 1051146 - Part 4: Introduce build system changes for the MOZ_WEBSPEECH_TEST_BACKEND flag. r=smaug, r=gps (16195e66e)
- Bug 1051146 - Part 5: Introduce new English model with no build integration. r=smaug (24996aa9c)
- Bug 1051146 - Part 6: Introduce build integration of new English model. r=smaug, r=gps (b718751a9)
- Bug 1051146 - Part 7: Introduce update scripts for Pocketsphinx and Sphinxbase code. r=smaug (5a85c7ed2)
- Bug 1051148 - Part 1: Patch that introduces WebSpeech API implementation. r=smaug (d18e5d86e)
- Bug 1171249 - Modified configure.in so MOZ_WEBSPEECH undefined implies both MOZ_WEBSPEECH_TEST_BACKEND and MOZ_WEBSPEECH_MODELS are undefined. r=gps (47d13fc61)
- Bug 1051148 - Part 2: Patch that turns off the WebSpeech API implementation using flags. r=gps (8b2ceef89)
- Bug 1051146 - Disable Pocketsphinx building on Windows due to bustage. r=gps (48f4a1c27)
- Bug 1204703 - Make nsContentSecurityManager scriptable (r=sicking) (fb514951a)
- Bug 1215745: Safer implemetation of nsJarChannel::AsyncOpen2. r=ckerschb (3ad0270b4)
- partial of Bug 1178892 - Split the profiler into Core & Gecko files and break apart monolithic headers. r=mstange (eda571469)
- Bug 1213407 - PluginModuleContentParent should delete its Transport. r=billm (153a54404)
- Bug 1169945 - Remove unused plugins.enumerable_names whitelist. r=bsmedberg (52126cbd8)
- Bug 1160166 - Disabled plugin meta information often isn't available to content processes. r=jimm (d63331192)
- Bug 1165981 - Only allow flash as a plugin on Windows 64-bit builds; r=bsmedberg (bfc073186)
- missing bit of 1164543 (5ef55bc55)
- Bug 1178963 part 1. Use nsIPluginTagInternal instead of nsPluginTag in cases where it may be nsFakePluginTag. r=peterv (f8702aead)
- Bug 1200484 (part 1) - Expose the column number in JS:escribeScriptedCaller(). r=luke. (85cf65a1c)
- Bug 1200484 (part 2) - Use JS column numbers in ReportWrapperDenial(). r=mrbkap. (1de0e6331)
- Bug 1200484 (part 3) - Use JS column numbers in WebSocket.cpp. r=baku. (3febdd8bc)
- Bug 1200484 (part 4) - Expose the column number in nsJSUtils::GetCallingLocation(). r=mrbkap. (479ae5e2a)
- Bug 1200484 (part 5) - Use JS column numbers in XPCNativeInterface::NewInstance(). r=mrbkap. (63e5e4e41)
- Bug 1200484 (part 6) - Use JS column numbers in ContentParent::OnChannelError(). r=billm. (4c84ef09d)
- Bug 1139718 - Warning when sending MM messages that include XPCOM data (r=smaug) (d50774823)
- Bug 1200484 (part 7) - Use JS column numbers in GetParamsForMessage() (bf19fc4cb)
- Bug 1200484 (part 8) - Use JS column numbers in EventSource. r=khuey. (64374ed83)
- Bug 1200484 (part 9) - Use JS column numbers in nsContentUtils::ReportToConsoleNonLocalized. r=khuey. (4d81c5f4f)
- Bug 1200484 (part 10) - Use JS column numbers in indexedDB. r=khuey. (64eec5da6)
- Bug 1200484 (part 11) - Set source location when initializing an nsJSScriptTimeoutHandler from a Function. r=peterv. (f135b3a2c)
- Bug 1200484 (part 12) - Use JS column numbers in nsJSTimeoutHandler. r=peterv. (8ec21dc04)
- Bug 1163664 - Don't check for plugin blocklist state on Android. r=jimm, r=jchen (14b8f3e57)
- Bug 1178963 part 2. Implement nsFakePluginTag to represent a non-NPAPI "plugin" that is actually implemented in JavaScript. r=peterv (ba11d6de0)
- Bug 1178896 - Add support for determining if a plugin is OOP to nsIPluginHost. r=aklotz (e8689641f)
- Bug 1178963 part 3. Make nsPluginHost aware of fake plugins. r=peterv (a2a9f8cb3) (ee118c4b8)
- partly import changes from tenfourfox:
- first rolling release: update to 91ESR EV and roots, HSTS, TZs, TLDs (d7c93b6cf)
- security: M1730935, M1735152, M1730048 (1603b00d0) (03efa0b17)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1170676: Fix null dereference in PluginModuleParent:treamCast; r=jimm (72e60e403)
- Bug 1164543 - Remove gAllInstances from PluginModuleChild (r=jimm) (15e035939)
- Bug 1164543 - Make plugin shutdown async in e10s (r=jimm) (e5239585f)
- Bug 1195527 - Part 1: Move D3D11 texture allocation into the TextureClient. r=jrmuizel (16e67d7e8)
- Bug 1195527 - Part 2: Add D3D11 video TextureClient recycler. r=jrmuizel (2c44d311b)
- more stuff todo (7b1bce288)
- Bug 1136945 - Add to Debugger.Memory docs possible outcomes of the nonincrementalReason in the onGarbageCollection event. r=fitzgen (33c6408ff)
- pointer style (05cc032b4)
- Bug 1197970 - Expose the GC cycle's number through Debugger.Memory's onGarbageCollection hook. r=sfink (5a420105f)
- Bug 1168992 - Part 1: Fix up reflection of classes. (r=Waldo) (5346e941b)
- Bug 1168992 - Part 2: Fix reflection of new.target. (r=Waldo) (38701b76d)
- Bug 1179063 - Hook up FunctionBox directly to the JSFunction being parsed to avoid allocating extra static scopes. (r=efaust) (d35252e18)
- Bug 1084319: Actually patter-match SIMD.bool (rebasing mistake); r=nbp (0980117b4)
- Bug 1200809 part 1 - Add InlinableNatives JitInfo infrastructure, use it for Math natives. r=bz,nbp (141e78aac)
- Bug 1193280 - Part 1: test stack trace handling of native frames with dedicated function. r=fitzgen (dd1c63891)
- Bug 1193280 - Part 2: Self-host Array.prototype.filter. r=jandem (82cbe1512)
- Bug 1200782 - Inline Array.isArray when we know the result statically. r=bhackett (546917507)
- Bug 1200809 part 2 - Convert Array natives to new InlinableNatives system. r=nbp (901aae723)
- Bug 1200809 part 3 - Convert String/RegExp natives to new InlinableNatives system. r=nbp (8e8eb68fd)
- Bug 1200809 part 4 - Convert Atomics natives to new InlinableNatives system. r=lth (430d8a6b8) (96d4680a6)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1190733 - Test initializedLength() instead of length() during the fast path for reversing unboxed arrays, r=jandem. (b5dcbd0e3)
- Bug 1070767 - Enable {Array, %TypedArray%}.prototype.includes in all builds. r=lth (de595c002)
- Bug 1195298 - Fix NewDenseArray intrinsic to work when the first argument is a double. r=till (1f551ada2)
- Bug 1190727 - Make initialization of temporary results array resilient against Array.prototype setters in self-hosted Map#next implementation. r=jandem (d705c623c)
- Bug 1200108 - Remove NewDenseArray intrinsic, use std_Array instead. r=till (e5c4126c6)
- Bug 1199822 - Turn self-hosting's cycle-check into an assertion; r=till (804600283)
- Bug 1194148 - Self-host Array.prototype.toString. r=till (4ffb4712c)
- pointer style (7b1a9900c)
- Bug 1200809 part 5 - Convert self-hosting intrinsics to new InlinableNatives system. r=till (8dd5eb0b9)
- pointer style (7974610a2)
- Bug 1200809 part 6 - Convert various natives to new InlinableNatives system. r=nbp (81d75199c)
- Bug 1200809 part 7 - Convert SIMD natives to new InlinableNatives system. r=nbp (5e67097e0)
- Bug 1200809 part 8 - Don't call shouldAbortOnPreliminaryGroups if we have an uninlinable native. r=bhackett (bd20f201f)
- Bug 1114507 - Part 1: Add/release the appId's refcnt if frame is in main process. r=kanru (741889791)
- Bug 1190903 - Don't send StopIMEStateManagement message after TabParent has been destroyed (r=masayuki) (97bba211b)
- Bug 1166592 - Remove ParentIdleListener from idle service when ActorDestroy() to avoid leaking ContentParent. r=khuey (771549a18)
- Bug 1114507 - Part 3: Remove PContetBridge channel when grandchild-process is killed. r=kanru (a84f888b3)
- Bug 1114507 - Part 2: Add/release the appId's refcnt in oop case. r=kanru (1fdb788b0)
- Bug 1114507 - Part 4: Test cases. r=kanru (889a770c7)
- pointer style (80bd2082b)
- Bug 1159347 - Make BaseProxyHandler::getPropertyDescriptor not-pure virtual. r=efaust (56de51919)
- Bug 1166847 - Implement OpaqueCrossCompartmentWrapper;r=evilpies (d762e785e) (fed708408)
- network: fix temporary variable type, this should fix not able to access port numbers that is larger than 32767. also updated bad port list from tenfourfox. (00fee1e0a)

* 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

@Roytam1 thanks for all your effort to maintain Kmeleon alive!

Roytam1 thank you very much for your weekly updates

Quote
cheetos1897
Crash would not happen if u Only visit the homepage of wikimedia.com. Try some content page.
such as this one
https://en.wikipedia.org/wiki/Olympic_Games

Sometimes KM doesnt crash after the page of wikimedia.com is completly loaded
but if u scoll the page up and down for several times it crashes rightaway.

Tried some recent Builds and it turns out that this issue starts to show since Biuld 20211009

Yes, same issue I have faced with the recent builds of KM Goanna, KM just closes when, sometimes, loads a new site or page, sometimes as soon I scroll down. After starting KM again it displayed the page sometimes without closing itself again.
This happens , i guess, on 1 out of 10 pages or sites.
Note. The last stable version, as far as in my experience, is KMGoanna build 2020-12-12, and This is the main browser on my system.



Edited 1 time(s). Last edit at 11/06/2021 06:57PM by WasiFH.

Quote
WasiFH

Quote
cheetos1897
Crash would not happen if u Only visit the homepage of wikimedia.com. Try some content page.
such as this one https://en.wikipedia.org/wiki/Olympic_Games
Sometimes KM doesnt crash after the page of wikimedia.com is completly loaded
but if u scoll the page up and down for several times it crashes rightaway.
Tried some recent Builds and it turns out that this issue starts to show since Biuld 20211009

Yes, same issue I have faced with the recent builds of KM Goanna, KM just closes when, sometimes, loads a new site or page, sometimes as soon I scroll down. After starting KM again it displayed the page sometimes without closing itself again. This happens , i guess, on 1 out of 10 pages or sites. Note. The last stable version, as far as in my experience, is KMGoanna build 2020-12-12, and This is the main browser on my system.

Have you found how to reproduce the problem? It should be very interesting in order to search for a fix.



Edited 1 time(s). Last edit at 11/08/2021 09:41AM by J.G..

@roytam1 the latest version of Malwarebytes gives two threads,
I guess they are false positive, however just see please:

Malware.Heuristic.1008, C:\USERS\*\KMGOANNA\ACCESSIBLEMARSHAL.DLL
Malware.Heuristic.1001, C:\USERS\*\KMGOANNA\IA2MARSHAL.DLL



Edited 1 time(s). Last edit at 11/09/2021 10:00AM by J.G..

Cant load pages on zhihu.com
try this url
https://www.zhihu.com/question/22540456

I notice that this browser has reached, theoretically, a very advanced version (76.4.5) and from a logical point of view I think it should be at about the same level as the browsers Google Chrome 76, Opera 76 or Firefox 76. Despite this huge number of the version, the way it behaves regarding the difficulty of having to access numerous sites and the problems that appear when try to loggin in make me think of the old Internet Explorer...

Quote
googleistheboss
I notice that this browser has reached, theoretically, a very advanced version (76.4.5) and from a logical point of view I think it should be at about the same level as the browsers Google Chrome 76, Opera 76 or Firefox 76. Despite this huge number of the version, the way it behaves regarding the difficulty of having to access numerous sites and the problems that appear when try to loggin in make me think of the old Internet Explorer...

You already know that the versioning numbers in K-meleon has nothing to do with those by the mainstream browsers. So are anything but at the same level of those examples. The K-meleon major version, 76, is static by one reason that has nothing to do with that.

In regard the K-meleon Goanna Build, it is using the engine version equivalent to Palemoon 27-28.

As a final note, no one here forces anyone to use the browser. If anyone feels that is not meeting their expectations, that person is free to use alternatives, but throwing shit over the engine assemble made by roytam, with the knowledge you have, it is unnecessary. Refer to the first sentence of this paragraph.

I think it occurs on some sites when page is being scrolled. Next time i will try to post a reproducible link or so, but only if i get some time.
Actually when it occurs, 2 or more times, i simply close KMG and use MyPal.
NOTE this is in reply to JGs post.



Edited 1 time(s). Last edit at 11/10/2021 03:59PM by WasiFH.

Just a comparison I don't understand why the older Goanna is better than the newer.

1. K-Meleon Goanna build 2020-12-12




1. K-Meleon Goanna build 2021-10-30




3. MyPal browser