Leaderboard

I think the problem is the way you usually call up the Thorium browser, or the way you open links in this browser. If you start the browser by applying the THORIUM_PORTABLE.bat file, a link to be opened cannot be transferred to this starter batch file via command line. Even if it is compiled to an exe file. I have tested that. If you start the browser by applying the thorium.exe file, a link can be transferred via command line but then all your flags inside the THORIUM_PORTABLE.bat file will be of course ignored. So, you have to add to those registry settings, where the thorium.exe file is called up, all your command line flags from the THORIUM_PORTABLE.bat file. This problem doesn't exist when calling up 360Chrome. You can successfully start this browser with a website transferred to its starter app 360Loader.exe via command line as a parameter. Try this: 360Loader.exe www.google.de And you will see, it works. And now try this: thorium.exe www.google.de You'll see, it works, too, but all your settings are not applied. However, when you try this: THORIUM_PORTABLE.bat www.google.de then you'll notice that it doesn't work. The browser will be called up but the Google website won't be opened. But when trying this: thorium.exe --user-data-dir="PathToYourUserProfile" --your-flags --single-argument www.google.de You'll see it works. The string PathToYourUserProfile has to be replaced by the complete path to your user profile and the string --your-flags by all your command line flags from the THORIUM_PORTABLE.bat file. The --single-argument flag is the one from the registry settings. Have a look at the files HKCR.reg and HKLM.reg inside the Thorium folder!

Getting a BOSD without a minidump is very rare. Do you really have the creation of minidumps enabled in your system? If so, maybe, the crash was so fast that the minidump couldn't be saved. But that's just an assumption. Besides this issue, we shouldn't forget that Thorium is still under development. I'm not surprised at all that BSODs happen, I'm literally expecting them. That's why Thorium wouldn't become a default browser in my Windows XP system at the moment. There are better alternatives for this purpose.

StartAllBack 3.8.2

Thanks, I will maybe consider your suggestion later on, now I bit pre-occupied with real life, I'm looking for a simple solution like modifying headers on the fly and make them switchable from presets.

Thanks for your opinion, unfortunately people wouldn't agree, and one the cases is documented here. probably you didn't see it yet. https://github.com/win32ss/supermium/issues/779

Can I request @Dave-Hto move the posts regarding CH spoofing, starting from this one: EDIT, and make all of the below to appear at the beginning, members already started to reply to the new topic. https://msfn.org/board/topic/186133-thorium/?do=findComment&comment=1270289 And then, move these posts, in this order. https://msfn.org/board/topic/186133-thorium/?do=findComment&comment=1270224 https://msfn.org/board/topic/186133-thorium/?do=findComment&comment=1270235 https://msfn.org/board/topic/186133-thorium/?do=findComment&comment=1270228 https://msfn.org/board/topic/186133-thorium/?do=findComment&comment=1270230 https://msfn.org/board/topic/186133-thorium/?do=findComment&comment=1270235 https://msfn.org/board/topic/186133-thorium/?do=findComment&comment=1270236 https://msfn.org/board/topic/186133-thorium/?do=findComment&comment=1270238 To the new, dedicated topic. Thank you! https://msfn.org/board/topic/186401-the-wretched-chrome-client-hints-another-doomsday-of-privacy-ways-out-of-it/

Sounds very much like Alienware might be blocking the switching process between the GPU. Give me the device ID.

On some visits there is a delay on load of a JS that can cause some issues if you try to use the site too fast. Such as first visit, I have to wait before I can look at notifications because if I click that too soon, I go to the notifications tab, rather than getting the toast on the index. Sometimes when viewing a page, it appears loaded but isn't. As you can tell from the small address bar that shows up in the bottom or there is a loading animation instead of the favicon in the browser tab. This script delay seems to be random to the user, heavy users will see it more often. But if this happens when you attempt a reply, then it can cause issues where the client looks to be in process of submitting but hasn't actually sent the POST method, or sometimes it has and the post is created but the client didn't refresh, or it just sits there forever but doesn't time out. This issue seems more prevalent on browsers that are customized with some sort of privacy. For example. this doesn't happen running on Iron with no extensions running, but trying to run the site in Firefox with any security extensions is an issue. I think that some of the external script sources are themselves using external sources, an issue I've seen on other sites and this cause a partial slow load on some sites.

New issue with Thorium, got a BSOD today without minidupms, The browser was open without any sites for quite some time, Several hours, maybe. Then I heard two short beeps, saw a short splash with BSOD, the PC went into reboot, Yes, on that PC, I do have the reboot option checked.

It's because they were boiling the frog very slow, now it's used up to the much wider extent.

Any user, who is going to join, needs to be warned in advance that he will be forced to adhere to the laws of the welcoming entity. Does github practise such policies? If no warning or demand follows, we could only give a flying frenkel. https://byjus.com/chemistry/frenkel-defect/

Me? No, I don't want anything from them, I'm not a member there, I only join Western European websites (like MSFN, and it's located in France). I adhere to the laws of my country, it is sufficient for me.

Sandboxie is a must-have utility with many security and non-security uses. This topic will focus on the compatibility and the uses of Sandboxie under WinXP. Postings about Sandboxie under other older operating systems (e.g. Windows Vista) are also welcome. Download links: 1) Windows XP 32bit (SP2 and SP3), SSE2 Sandboxie v5.40 (WinXP, SSE2, 9Apr2020) is the last version for Windows XP 32bit (SP2 and SP3), SSE2. It can be downloaded from: https://github.com/sandboxie-plus/Sandboxie/releases/download/v5.40/SandboxieInstall32_xp.exe v5.40 is freeware and was created by David Xanatos, based on the source code of v5.40 released by Sophos. v5.40 does not require registration or activation. v5.40 (9Apr2020) is of special interest because many Windows XP users may think that v5.22 (30Oct2017) is the last version of Sandboxie for Windows XP, maybe based on the info at Wikipedia: "Windows XP SP3 and Windows Vista SP2 were supported [=by Sophos] up to version 5.22, after which their support [=by Sophos] was dropped." https://en.wikipedia.org/wiki/Sandboxie The subsequent developer David Xanatos, however, had created in April 2020 the above special edition v5.40 for Windows XP, which is a little gem. v5.40 by David Xanatos works great under Windows XP (32-bit) on computers with an SSE2 CPU (= Pentium 4 and higher). Do NOT attempt to install v5.40 on an SSE-only CPU (e.g. Pentium 3). 2) Windows XP SP2 and SP3, SSE-only Sandboxie v5.22 (30Oct2017) is the last version for Windows XP SP2 and SP3, SSE-only (e.g. Pentium 3). 32-bit installer: https://web.archive.org/web/20180107162521if_/https://www.sandboxie.com/attic/SandboxieInstall32-522.exe 32+64-bit-installer [but not for Windows XP x64]: https://web.archive.org/web/20171109032037if_/https://www.sandboxie.com/SandboxieInstall.exe 64-bit installer: https://web.archive.org/web/20180107162355if_/https://www.sandboxie.com/attic/SandboxieInstall64-522.exe v5.22 is a limited trial version by Sophos, with the following limitations of the trial version: - it does not expire, but a nag screen will appear after 30 days - you may create several sandboxes, but you cannot use several sandboxes at the same time - you cannot use the Forced Programs and Forced Folders features - registration was for a time-limited subscription, requiring re-activation - you cannot register or activate v5.22 at Sophos anymore

Hi @Tihiy I just checked version 3.8.2 and the problem I reported earlier regarding EA app is gone so, good job. Sadly I found a new problem with another program. To the point... After the installation of Auslogics Disk Defrag (version 8.0.24 - This is an older version of this program, yes I am aware of that but I use it because it is the last version that has full functions not hidden under unfair paid license) an empty icon remains on the taskbar. There was no such thing at all in version 3.7.11 of StartAllBack and this icon simply disappeared after the installation was complete. This icon disappears when you click on it. However, the icon does not remain on the taskbar at all when the program is installed with a different compatibility e.g. Windows 8 or 7. The problem also affects the program's uninstaller, which also leaves an empty icon after uninstallation, unless its compatibility is also changed before to Windows 8 or 7. I know that this may not be a super important issue but I thought it might be worth mentioning because it might help prevent similar problems from occurring with other programs. Screenshot of the issue.

The START menu still behaves strangely when I click the mouse until I press the WIN key, then the START menu appears correctly. W22H2 SAB 3.8.1 Video_2024-08-06_171905.mp4

My findings to add: By default Sandboxie driver (HKLM/Sys/CurrCtrlSet/Services/sbiedrv) registry entry is flawed from way back when v4.20 was released - instead of "DependsOnService" it should have "DependOnService" (FltMgr) - I have contacted David Xanatos about it and he acknowlegded my observation. The second thing was observed in installing XP with harkaz's USP4 - again in registry - Control/GroupOrderList: for "Filter" group tags should go up to "0b" and not only "06", so 07 00 00 00, 08..., 09..., 0a..., 0b... should be added to that group or Sandboxie (at least 5.22) complains it cannot hook in its driver at system boot (with other minifilters installed, such as diskeeper's or perfectdisk's). Tags (component load order monikers) are very important especially in older Windows systems - even if services containing them are not loaded. The other group lacking mention in GroupOrderList is SCSI Class - should go up to 2d 00 00 00 because there is one driver containing such a Tag. While it may seem superficial, when Tag-problems are corrected, the whole system is much more responsive and also faster to boot. This way Sandboxie functions very well in my install of XP_USP4 with Avast, Privatefirewall, SBie and Diskeeper12 (all programs that use low-level drivers with XP_SP3's limited fltmgr).

@AstroSkipper Just check this out - Office 2007 UpSourcer. Real problem solver There is also 2010 ones which is realy builded to works with XP and 2003 too! https://forums.mydigitallife.net/threads/upsourcer-slipstream-patches-for-office-2007-2010-2013-2016-and-rebuild-setup-source-files.87111/#post-1794847

Sandboxie - Compatibility (operating systems and CPU) The following are the last versions of Sandboxie working OK under: Windows XP SP2 and SP3 - SSE2 CPU (Pentium 4, AMD Athlon 64, Intel Atom, all x64-capable CPUs, and later) Sandboxie v5.40 (9Apr2020) Windows XP SP2 and SP3 - SSE-only CPU (Pentium 3) Sandboxie v5.22 (30Oct2017) Windows XP x64 - ?? Windows Vista 32bit - ?? Windows Vista 64bit - ?? It is unclear which version of Sandboxie is the last/best version to run under Windows Vista.

The "portable" r3dfox is NOT portable. Left crap in my AppData directory and in my registry. The LibreWolf "portable" appears to indeed be PORTABLE. Have not found anything "left behind" by it. That wins applause in my book. But not the telemetry stuff that seems basically no different than "regular" Firefox.

119 with second launch after turning off the start page "shortcuts". Again pretty much all the same domains as "regular" Firefox.

165 connections with r3dfox first launch. But clearly several of them are with the "default" start page that you don't even know is there until after you perform that first launch. Twitter, Facebook, Amazon, Reddit, YouTube, and Wikipedia have all been "notified" that I just installed r3dfox

Third Firefox launch. 212 DNS Connections without ever opening a single solitary tab. And uBO doesn't "reveal" ANY OF THEM! If we are truly talking about "privacy concerns", by all means, let's do so, but let's not turn a blind eye to the TONS of TELEMETRY that Firefox performs without most users ever knowing it is happening!

All I can tell you is that Firefox is a TELEMETRY NIGHTMARE in comparison to UNGOOGLED CHROMIUM. My very first run of a just-downloaded Firefox on a computer that has NEVER had Firefox on it showed 340+ DNS ENTRIES UPON FIRST RUN !!! THREE HUNDRED AND FORTY PLUS !!! Disable everything I can find for the "start page" and I still show 153 DNS ENTRIES WHEN I LAUNCH FIREFOX !!! ONE HUNDRED FIFTY THREE !!! Without ever opening a single solitary tab, that's just to launch Firefox !!! I can launch my Ungoogled Chrome and let it sit for HOURS upon HOURS and it will NEVER MAKE A DNS CONNECTION if no tabs are ever opened.

Done (I hope!) I can't do anything about the order the posts appear in. The forum software just puts them in strict chronological order.

@Cixert Here are the sizes and number of files more detailed which were downloaded in my WSUS download sessions including the updates I manually added as their download failed: Windows XP x86: Size and number of files in the folder client\cpp: 87.6 MB and 12 files Size and number of files in the folder client\dotnet: 286 MB and 3 files Size and number of files in the folder client\dotnet\x86-glb: 673 MB and 95 files Size and number of files in the folder client\msse\x86-glb: 115 MB and 4 files Size and number of files in the folder client\wddefs\x86-glb: 55.4 MB and 1 file Size and number of files in the folder client\win\deu: 33.1 MB and 19 files Size and number of files in the folder client\win\glb: 115 MB and 4 files Size and number of files in the folder client\wxp\deu: 725 MB and 310 files Size and number of files in the folder client\wxp\glb: 10.0 MB and 3 files Office 2007: Size and number of files in the folder client\o2k7\deu: 430 MB and 3 files Size and number of files in the folder client\o2k7\glb: 31.9 MB and 2 files Size and number of files in the folder client\ofc\deu: 81.8 MB and 7 files Size and number of files in the folder client\ofc\glb: 5.08 GB and 572 files Office 2010: Size and number of files in the folder client\o2k10\deu: 628 MB and 1 file Size and number of files in the folder client\o2k10\glb: 101 MB and 1 file Size and number of files in the folder client\ofc\deu: 19.7 MB and 1 file Size and number of files in the folder client\ofc\glb: 5.00 GB and 200 files @Cixert Please copy this list, replace my values with yours for easier comparison and post it here! BTW, I have always enabled the option for the download of Service Packs. That's probably the reason for the different size of the folder client\wxp compared to yours. And it should be mentioned that Office 2010 is not XP-compatible as this topic here is located in the Windows XP subforum. That's why I downloaded Office 2010 updates with WSUS Offline Update 11.9.6 instead of WSUS Offline Update 9.2.6, BTW. In any case, it is only included here for reasons of comparison regarding downloading updates with WSUS Offline Update in 2024. Greetings, AstroSkipper

The downloads depends on the weather. That's what my grandmother would have said if she were still alive. You are simply great.

I didn't say that the files in the client\ofc\glb folder are exclusively for Office 2010 or higher. I said: Inside this folder, there are of course many updates for Office 2007 (12.0.xxxx.xxxx), too. Simply have a look at the xml files inside the cab files! All files with the version number 12.0.xxxx.xxxx are targeting Office 2007. As you already cited the creators of WSUS Offline Update 9.2.6 correctly, the download of updates for older Office versions leads to dynamically determined updates which are also downloaded. Either leave them in that folder or delete them before creating the iso image!

@Cixert Just for testing purposes, I have downloaded all Office 2007 updates again but this time under Windows 10. For this, I used WSUS Offline Update 9.2.6 and the wsusscn2.cab file from 16.07.2018 via Wayback Machine https://web.archive.org/web/20180716122029/http://download.windowsupdate.com/microsoftupdate/v6/wsusscan/wsusscn2.cab due to your hint about End of updates in April 2018. And what shall I say? It is still possible. I checked the generated download log, and I compared the downloded files with my old download archive. Only a few downloads failed which can easily be found via Wayback Machine and added. Here is a list of failed downloads: The folder o2k7 wasn't created at all as only these 5 files, whose downloads failed, had to be downloaded into this folder. Therefore, I created it and its subfolders, and then copied all files into. All downloaded files seem to be related to Office. However, I can confirm there are many updates downloaded for higher versions than 12.0.xxxx.xxxx which are not targeting Office 2007 but Office 2010 (14.0.xxxx.xxxx), Office 2013 (15.0.xxxx.xxxx) and so on, all inside the folder client\ofc\glb. So, your cited statement seems to be correct: But these files could be of course deleted if unwanted as the folder client\ofc\glb only contains 570 - 580 files depending on the used wsusscn2.cab version. It should also be mentioned that I have added the following exclusions to the ExcludeList-ofc.txt file as a precaution: Anyway! I would suggest to redownload all Office 2007 updates under the above listed conditions. Cheers, AstroSkipper

On a Win10 PC?! Why would you need XP browsers, then? If you're a die-hard of China issued browsers, I suggest this one. Minimum telemetry, very fast! Works on Win7+. Dave, sorry for the slight off-top. https://msfn.org/board/topic/186346-i-moved-to-centbrowser-i-feel-happy-about-it/ https://static.centbrowser.com/win_stable/5.1.1130.129/centbrowser_5.1.1130.129_x64_portable.exe Be Warned Some Telemetry is still there.

Spam flood fixed.

Here are my first results: Size of the folder client\o2k10: 729 MB Size of the folder client\ofc: 5.02 GB As already mentioned, I used WSUS Offline Update 11.9.6. My download log doesn't show any failed downloads. So what sizes did you get? Shall we also compare the number of files in each folder?

As far as I could see, some updates are missing. Now, I am downloading again but using the wsusscn2.cab file from 11.02.2021 which fits in with the fact that support of Office 2010 ended on 13.10.2020.

Self-evident is another very important matter, and I quote. "2. We have the right to refuse or remove any User-Generated Content that, in our sole discretion, violates any laws or GitHub terms or policies. User-Generated Content displayed on GitHub Mobile may be subject to mobile app stores' additional terms." In light of the fact I posted yesterday, about the Russian government that has officially legalised intellectual property theft and made it a law. In the quote above, Github declares it may act on any laws, not only US laws, so it may allow Russians to steal members' intellectual property, and those members can't do anything about it, they had been warned, and accepted github terms. For example, Russians write to github and demand them to open up someone's work, or take it down. But if a Russian member hosts stolen property on github, it's all according to their "law", so all is "fine". Sources: https://docs.github.com/en/site-policy/github-terms/github-terms-of-service https://ipwatchdog.com/2022/03/16/russian-businesses-can-now-legally-steal-intellectual-property-unfriendly-countries/id=147528/

Please don't place Canadian laws in one row with those, it's no good to lump them altogether.

And here is my screenshot with a 2004 game Sacred Plus. It only supports CSAA, but not MSAA. Which is not surprising at all.

I think they do ignore, and the project gone due to the tiredness of the developer, but then again, it's a speculation since we saw no statement, even though he is a current member.

Hi @TheLeftOldComputer! Many antimalware programmes are no longer compatible with Windows XP but ClamWin has always been. Unfortunately, it does not have a real-time protection. Only an additional tool called Clam Sentinel which has no longer been developed for years and never was a good one. It could only use ClamWin to scan downloaded files. So what about a programme based on the open source ClamAV with its virus definition databases which offers real-time protection? Or a fork of ClamWin with real-time protection and still compatible with CPU's only having the instruction set SSE? Kind regards, AstroSkipper

Wow! I knew it! This new member now posts fake programmes. The account with the nick @lll888. I'd like to notify @Tripredacus. It's already quite a lot of posts that member made with dubious software and batch scripts, now this link to fake Kernel.

Thank you very much for the investigation, I felt like something fishy was going on there. I'll upvote tomorrow (I'm out of likes, they are limited).

I made a lot of posts on Windows 8.1 and software compatibility. In terms of Firefox, all you need to do is to lower the system requirements of Firefox.exe with CFF Explorer from 000A to 0006. There are no additional functionalties needed. So it still natively supported on Win8+ Office 365 (the latest beta 2407) works on 8.1 with an edit to Registry. No compatibility issues. I hope the compatibility won't be broken till the end of this year, so that consumer versions of the upcoming Office 2024 can also be installed on Windows 8.1 (they will be supported till 2029) Chromium based apps tend to create bigger issues on 8.1

The Pegasus NES was sold in Poland. Dendy system were more of a thing people knew existed but didn't know about, more popular to know about than the Pegasus. I wouldn't consider Russia to even be a country to use as an example for famiclone proliferation, rather Hong Kong was the primary source for those. On translations, "just aware" meaning that I've heard of it before, not to mean that I just heard of it from this link. One of my side-projects is research into international release of computers and video games. Gruz is Polish word, not Russian, as one would expect that a book written in Polish would be using their own words. It translates to "rubble" as per Google translate, so I wonder if it is to mean "junk" like "retrojunk" type thing people from English speaking countries may refer to something as. Maybe a Polish speaking user can post here what the title is supposed to mean.

but unfortunately we(including me, TFF, MCP, WF, SM, Moz, Google, Apple, etc.) are chasing the same rabbit.

yeah because these websites trends to use lots of events and workers in background, that is the thing I can't have much to do with. hope I can have some more time to port some UXP features to moebius, but don't hold your breath on it.

New regular/weekly KM-Goanna release: https://o.rthost.win/kmeleon/KM76.4.5-Goanna-20211127.7z Changelog: Out-of-tree changes: * update Goanna3 to git 20d9ef889...8cdfd3201: - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1170039: Fix printf format specifiers in GC stats. r=terrence (739a9203f) - Bug 1169222 - Add --debugger option to jittests to run tests under the specified debugger r=terrence (7d7d6f93b) - Bug 1170596 - JitSpewer: Always flush after dumping the graph to the files. r=bbouvier (cabece9da) - Bug 1018628 - Part 3: Add test for default parameter with destructuring default value. r=jorendorff (923a193ea) - Bug 1142816 - js/src/jsapi-tests/testGCHeapPostBarriers.cpp has leaks. r=erahm (cc543a475) - Bug 1252075 - use UniquePtr instead of ScopedDeletePtr in testGCHeapPostBarriers; r=terrence (f58ddfed2) - Bug 1168293 - Remove MOZ_GLUE_IN_PROGRAM from config/config.mk. r=mshal (93877bb3a) - Bug 1179695 - Remove mWasAsync from WidgetQueryContentEvent. r=masayuki,nchen (734c1a93f) - Bug 1185316 TabParent should flush pending IME notifications after child process handles all sending events r=smaug (d2eb7c359) - Bug 1153586 - Prevent the same BUTTON_DOWN event from opening the dropdown menu again on platforms that don't consume the event. r=enndeakin (5b8bf9156) - No bug - Add a bug number to the historical.html expectations. (9377c1ff6) - Bug 1161950 - Remove support for createEvent("CloseEvent") / initCloseEvent, r=smaug (9f2e5b970) - Bug 1161950 - patch 2 - fix the webplatform test, r=Ms2ger (04046e15a) - Bug 1171294 - Stop warning if piTarget is null. r=smaug (5ab612a71) - Bug 873923, add a method to allow popups to be anchored at a rectangle, r=tn,mark,khuey (08b7902c9) - Bug 873923, anchored at popup test, r=test (36bb7091a) - Bug 250091, support page up and page down in menulists, r=neil (6e455d5c6) (78b33706e) - ported change from `dev' branch of rmottola/Arctic-Fox: Bug 1167607 - Alter behavior of AsyncEventDispatcher for dispatching chrome only event on chrome node. r=smaug (7c7bf799c) - import changes from `dev' branch of rmottola/Arctic-Fox: - fix patch ordering coming from PM repo (3a8768f44) - Bug 1119074 - If we're stopping event propagation from XUL popups, also stop them from crossing process boundaries r=smaug,felipe (965e2193e) - Bug 1082145 - |js::WatchGuts| can leak |wpmap|. r=erahm (12c86f3d3) - No Bug - Improve Interpreter stack accessor assertions. (rs=Waldo) (8ba7a702c) - Bug 874842 - Return Event instead of nsIDOMEvent (748b57fd2) - Bug 1165966 - Add test cases r=terrence (79a909d5b) - Bug 1167025: Do not mix bool and int in bitwise-or in blendpsMask; r=sunfish (f0f23c0c4) - Bug 1158323 - Make sure we set a base rect on document elements that have margins set. r=tn (c253a2ef8) - Bug 1178847 - Move the code from ChromeProcessController::InitializeRoot to APZCCallbackHelper so it can be reused in the child process. r=botond (16d539bcb) - Bug 1165966 - Add error checking when populating safepoints r=bhackett (c66d249d1) - spacing and pointer style (cddc1bac4) - Bug 1196027 - check the actual current marking mode instead of the permanent intention, r=terrence (eddcfd7fb) - Bug 1206590: Move gcWeakMapList from JSCompartment to JS::Zone. r=terrence (7e5e0d505) - Bug 1181908 part 1. Fix support for JSOP_OBJECT in scripts parsed on background threads by clearing the unboxedLayouts list on the background thread parsing compartment when merging the parse result to the target compartment. r=jandem (25c6a3b01) - Bug 1163207 - Make RematerializedFrame store the real callee. (r=shu) (ce276e91c) - Bug 1164448 - Handle unwound rectifier frames as exit frames in JitProfilingFrameIterator. r=jandem (bb639b4e2) - Bug 1164448 - Add test. r=jandem (83f5cc608) - Bug 1196497 - Don't assert that the replacer continues to pass IsArray during JSON.stringify. (If the replacer was a revocable proxy to an array, revoking the proxy would make the replacer no longer IsArray.) r=evilpie (442c3823f) - Bug 1177247 - Prevent HandlePossibleViewportChange from clobbering a restored scroll position from forward/back navigation. r=botond (4202ac757) - Bug 1182772, optimize ProcessGlobal out from CC graph (and also TabChild's EventListeners), r=mccr8 (ccb2278bf) - Bug 1139155 - Add a mechanism to know when the APZ is done processing. r=botond (17328e5be) - Bug 1171537 - Allow URIs to be the empty string in TabParent::RecvCreateWindow. r=billm. (e280e994c) - Bug 1173219 - Return nsresults from TabParent::RecvCreateWindow to make opening windows more robust. r=billm (9f0633b15) - Bug 1142817 - Use UniquePtr in testXDR_sourceMap. r=erahm (7ec437162) (7a25ca546) - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1172785 - Switch to ECDSA for MTI suites, r=ekr (bfb0d4cd3) - Bug 8607346 - Add -fvisibility-inlines-hidden to VISIBILITY_FLAGS to hide inline member functions. r=glandium (e5604fe00) - Bug 1119072: Remove outdated comment. DONTBUILD (386a2eb0f) - Bug 1142826 - js/src/jsapi-tests/testMutedErrors.cpp has leaks. r=luke (b980ccb65) - Bug 1164101 - Avoid unreachable code warning in shell.js; r=bc (7d7faa019) - Bug 1142820 - js/src/jsapi-tests/testPersistentRooted.cpp has leaks. r=jonco (7fcc2e586) - Bug 1160182 - Handle extended functions in CreateLazyScriptsForCompartment. r=shu (5b51e9120) - Bug 1145785 - Fix OOM crash in the ARM simulator's icache check. r=bhackett (db37ce2b0) - Bug 1139856 - do not lose wakeups. r=luke (e10f7a491) - Bug 1160146: Disable global sized deallocation for VS2015. r=glandium (353923945) - Bug 1155965 - Part 2: Verify that CreateOffer/Answer has been called when SetLocal is called. r=drno (861c26dc7) - Bug 1155965 - Part 1: Test-cases. r=drno (3a00ef585) - Bug 1157766: mismatched DataChannel initial channel size in JSEP database breaks adding channels r=bwc (3071ab5ac) - Bug 1161136: Ensure that mid is copied over before we disable answer m-sections. r=jesup (97ef20403) - Bug 1096795 - Put a=rtcp in SDP when gathering ends. r=mt (8979c4634) - Bug 1164582 - Report an error when the appdata portion of a webrtc msid is missing. r=mt (4f0af4d90) - Bug 1164575 - Log SDP in SetLocalDescription. r=mt (c8685bab1) - Bug 1160280: Put ssrc attributes in recvonly m-sections. r=ekr (17db960d0) - Bug 911450: sndio webrtc audio backend, build integration bits r=jesup (68cad0300) (8cdfd3201) * 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

it may because it tries to detect a custom URL scheme with javascript to see if there is a program can deal with spotify "URL" and give a spotify download if there is none, and the mechanism it use may be blocked by latest security patches.

twitter creates 1500+ events per tab nowadays, that's the reason UXP browsers having hard time when browsing twitter.

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::SetNewDocument 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::XPCONNECT_SHUTDOWN. r=terrence (815cbd630) - Bug 1171722 - Part 3: Specify gcreason::XPCONNECT_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

New regular/weekly KM-Goanna release: https://o.rthost.win/kmeleon/KM76.4.5-Goanna-20211016.7z Changelog: Out-of-tree changes: * update Goanna3 to git fae40c10f...f355dbe3d: - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 774364 - Part 4: Add setRNGState testing function. r=sstangl (648a491f0) - Bug 1206591 - IonMonkey: MIPS32: Move LIRGeneratorMIPS::visitRandom to architecture-specific. r=arai (edcb0fa60) - Bug 1205917 - IonMonkey: MIPS32: Use rshiftPtrArithmetic in clampIntToUint8. r=arai (10e234d60) - Bug 1205921 - IonMonkey: MIPS32: Fix scratch register in MacroAssembler::memIntToValue. r=evilpie (5682d93f1) - Bug 1204604: Check for int32 value, not int32 type, in SIMD.T.{replaceLane,shuffle,swizzle}; r=lth (18dca9450) - Bug 1180763 - Part 1: Add the size of the tenured object to the tenure promotions log entry; r=sfink (456827a5e) - Bug 1180763 - Part 0: Document the size property of entries in the tenure promotions log; r=sfink (96244a8f9) - Bug 1180763 - Part 2: Test the size property of entries in the tenure promotions log; r=sfink (602225219) - Bug 1208867 - Add xxx64 operations to MacroAssembler-none. r=sstangl (c3a8207bf) - Bug 1203964 - Part 1: Move MacroAssembler::rshiftPtr into generic macro assembler. r=h4writer (2c6486a37) - Bug 1203964 - Part 2: Move MacroAssembler::rshiftPtrArithmetic into generic macro assembler. r=jandem (ce3e31ade) - Bug 1203964 - Part 3: Move MacroAssembler::rshift64 into generic macro assembler. r=hev (8f395fad1) - Bug 1203964 - Part 4: Move MacroAssembler::lshiftPtr into generic macro assembler. r=nbp2 (99a45b756) - Bug 1203964 - Part 5: Move MacroAssembler::lshift64 into generic macro assembler. r=sstangl (3bcd9eb5f) - Bug 1136806 - Fix assertion failure with Debugger and oomAfterAllocations; r=shu (efcab9ab7) (3d7e107cc) - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1149185 - override GetMessageLoop() r=nical (42bec9405) - Add a "blacklisted" property to d3d11 telemetry. (bug 1187453, r=mattwoodrow) (83bcb0e8e) - Don't create a D3D11 compositor device on content processes. (bug 1183910 part 8, r=bas,mattwoodrow) (c03d496a6) - Bug 1168889 - ContentClientSingleBuffered needs to upload during the transaction. r=nical (35171fe14) - Bug 1176077 - Only recompute PaintedLayer item visibility when display list has changed. r=mattwoodrow (0ea691594) - Bug 1179298. ClientTiledPaintedLayer should allow empty transactions to succeed if it doesn't need to draw anything. r=mattwoodrow (4aaad97b2) - Bug 1169200 - Send OpRemoveTextureAsync before actor destroyed r=nical (86007156c) - Bug 1158838 - Add some gfxCriticalError logging. r=milan (32bf80848) - Bug 1129871 - Avoid copying TileClient when not necessary. r=BenWa (1ac30bcc2) - Bug 1186911 - Fix progressive paint when using tiled-drawtarget. r=nical (87649fb8d) - Bug 1118876 - Add edge padding with DrawTargetTiled. r=jrmuizel (64b12cf6a) - Bug 1157677 - Fix uninitialized members warning in ClientTiledLayerBuffer. r=nical (d11d0da30) - Bug 1151071 - Make sure low-resolution tiles are transparent. r=mattwoodrow (164446378) - Bug 1170390 - Use Optimal2DFormatForContent over OptimalFormatForContent to avoid enum conversions. r=jrmuizel (db7ee2c31) - Bug 1180326 - Part 1: Abstract parts of the client side tiling that we need. r=jrmuizel (0ad7651fe) - Bug 1179987 - Remove painted region from SurfaceDescriptorTiles since it isn't used any more. r=nical (653e9389e) - Bug 1180326 - Part 2: Add support for variable tile sizes. r=jrmuizel (081349a5e) - Bug 1180326 - Part 3: Move repeated code into GetBackBuffer. r=jrmuizel (2e9091ae1) - Bug 1180326 - Part 4: Add new content client. r=jrmuizel (7362d805b) - Bug 1180326 - Part 5: Support HWC with tiling when we only have 1 tile. r=jrmuizel (09ce93b7e) - Bug 1152461 - Don't use tiled layers for overflow:hidden scroll frames. r=tn (804a683a1) - Bug 961887 - Add some reftest fuzz. (dcd707004) - Bug 1186061 patch 1 - Add feature to reftest harness to allow skipping the flush. r=mattwoodrow (8cefe9640) - Bug 1176969 - Disable OMT animation for any frame in a preserve-3d scene rather than only frames whose parent and child are in a preserve-3d scene. r=mattwoodrow (693d0a518) - Bug 1176969 followup - Annotate new reftests as intermittently fuzzy on Mac. (0269d5d8f) - Bug 1176969 followup - Increase error amounts for intermittent fuzzy-if(cocoaWidget). (116039ed0) - Bug 1176969 followup - Mark test as random for now until I have a chance to rewrite it. (8ba561f83) - Bug 1176969 followup - Reorganize the timing of the reftest to make it more reliable. (f3ffc93cd) - Bug 1186061 patch 2 - Disable compositor thread animation of transforms when backface-visibility is hidden. r=mattwoodrow (e67bed9fb) - Bug 1169666 - Mark reftests failing due to scrollbar transparency random on GTK3. r=karlt (0c6080d55) - Bug 1180326 - Part 6: Use SingleTiledContentClient for non-scrollable layers on b2g and OSX. r=jrmuizel (079951f0a) - Bug 1189261 - Mark tile invalid region in coordinates relative to the tile. r=jrmuizel (7e696c911) - Bug 1187619 - Only optmimize FrameLayerBuilder visibility calculations if correct. r=mattwoodrow (05228ad31) - Bug 1190950: Check mappings and surfaces for UpdateFromSurface more robustly. r=mattwoodrow (c00adba10) - Remove the backend flag to TextureClient::CreateForDrawing. (bug 1183910 part 9, r=mattwoodrow) (5df1d964c) - Bug 1135935 - Part 1: Don't implement ISurfaceAllocator for the texture recycler. r=sotaro (bcba93d3d) - Bug 1135935 - Part 2: Merge TextureClientRecycleAllocator into base class. r=sotaro (8f4e7b6b1) - Bug 1135935 - Part 3: Allow overriding of allocations for texture client recycling. r=sotaro (35d7499e9) - Bug 1135935 - Part 4: Move IDirect3DTexture9 allocations and ownership into the TextureClient. r=jrmuizel (e62cf9acf) - Bug 1135935 - Part 5: Add D3D9 texture recycler. r=jrmuizel (8ab85b064) - Bug 1135935 - Part 6: Wait for the compositor to stop using textures before recycling them. r=jrmuizel (ae0c45cfd) (96836f9b3) - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1156600 - asmjscache: fix assert when reading metadata file for cache write fails (r=janv) (2b9895e97) - Bug 1173756 - Part 1: Update asmjscache to use PBackground; r=luke (adb326d80) - Bug 1173756 - Part 2: Merge MainProcessRunnable and ParentProcessRunnable; r=luke (b42ffa5f3) - Bug 1173756 - Part 3: Merge File and ChildRunnable; r=luke (961701adf) - Bug 1194555 - Part 0: Cleanup GetReportsState constructor. r=njn (4ae530854) - Bug 1145050 - Fix uninitialized class members in Http2Session. r=hurley (a12a9d372) - Bug 1132357 - remove h2-draft* support. r=mcmanus (24b8f347f) - Bug 986302 - Add memory reporting for HPACK tables r=hurley r=njn (0d2ebae56) - Bug 1123386 - Part 1: Update the clang build script for the most recent clang; r=rail (394f24496) - Bug 1123386 - Part 3: Mark DoNothingCallback::Callback as override in order to fix a build issue in warnings as errors mode (1d236acbf) - Bug 1123386 - Part 2: Temporarily relax the libstdc++ symbol version requirements for Linux static analysis builds; r=glandium (1415a4d90) - Bug 828844 - Add a "system-heap-allocated" memory report on Linux. r=glandium. (eead43b4e) - Bug 1194061 - Implement "system-heap-allocated" reporter for Windows. r=dmajor. (228385cb4) (d0c58681f) - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1194555 - Part 1: Remove RunReportersForThisProcess. r=njn (5c745b3ce) - Bug 1145007 (part 1) - Test |residentUnique| in test_memoryReporters.xul. r=erahm. (bc2fa3d30) - Bug 1145007 (part 2) - Add a "resident-peak" distinguished amount and memory reporter on Unix. r=erahm. (007256c58) - Bug 1194555 - Part 2: Remove |explicit| attribute from nsIMemoryReporterManager. r=njn (a0db6068e) - Bug 1203786 - test_memoryReporters.xul should ignore namespaces. r=njn (04e8842aa) - Bug 1194555 - Part 3: Remove |getReportsForThisProcess| from the nsIMemoryReporterManager interface. r=njn (cc3282c7f) - Bug 1194555 - Part 4: Rename GetReportsState PendingProcessesState. r=njn (3dab1b600) - Bug 1194555 - Part 5: Add option for async memory reporters. r=njn (f05ae83da) - Bug 1194555 - Part 6: Run reporters asynchronously. r=njn,jld,ted (d513fcd77) (f355dbe3d) * 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

New regular/weekly KM-Goanna release: https://o.rthost.win/kmeleon/KM76.4.4-Goanna-20210918.7z Changelog: Out-of-tree changes: * update Goanna3 to git 8d7b56d38...9e8144b17: - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1129873 - [GTK3] Implement wrapper to GtkAppChooserDialog to allow using native application chooser. r=karlt (98efe5d61) - Fixup for bug 1129873 for bustage of GTK+3 builds on elm. r=karlt. DONTBUILD. (1470214b7) - Bug 1175394 part 1 - Create an unmapped (strict) arguments object for functions with rest/destructuring/default arguments. r=jorendorff (59eebb140) - pointer style (4eefdc553) - Bug 1151149 - Throw if either the target or handler is a revoked proxy in Proxy constructor. r=efaust (c23d01862) - Bug 1158222 - Implement the new [[Set]] receiver restrictions. r=jorendorff (fc4fbc164) - Bug 1049662 - Update ES6 scripted proxies ownKeys trap to ES6 final. r=efaust (dd8eb6240) (4a7fbafa2) - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1178033 - XDRBuffer: Replace isUint32Overflow by a simple check of the capacity max. r=luke (522f39612) - Bug 1191543 - Remove AutoNameVector and replace with Rooted; r=jonco (979b2fe42) - pointer style (cf34bfcf1) (45e0dadc1) - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1181443 (part 3, attempt 2) - Remove nsTHashtable::EnumerateEntries(), which is no longer used. r=froydnj. (3368831d9) - Bug 1183141 - Use nsTHashtable::Iterator in ServiceWorkerManager::ClaimClients; r=catalinb (6ec0099ba) - Bug 1182966 - Part 1: Use nsTHashTable::Iterator in DomainSet::CloneSet. r=mrbkap (1c03a23d9) - Bug 1182966 - Part 2: Stop referring to the nsI class for no apparent reason in DomainPolicy. r=mrbkap (67c9e70a2) - Bug 1182977 - Use PLDHashTable::Iterator in dom/storage/. r=khuey. (accb68cb3) - Bug 1182977 (part 2) - Use PLDHashTable::Iterator in dom/storage/. r=khuey. (0dcf0792f) - Bug 1182959 (part 1) - Use nsTHashtable::Iterator in CertBlockList. r=honzab. (165e4d2d4) - Bug 1182959 (part 2) - Use nsTHashtable::Iterator in CertBlockList. r=honzab. (d24e98945) - Bug 1182959 (part 3) - Use nsTHashtable::Iterator in CertBlockList. r=honzab. (a5c240594) - Bug 1182959 (part 4) - Remove BlocklistSaveInfo. r=honzab. (2e11db520) - Bug 1182959 (part 5) - Use nsTHashtable::Iterator in nsCertOverrideService. r=honzab. (f560b8677) - Bug 1121760 (part 4) - Remove PL_DHashTableRawRemove(). r=poiru. (33267404b) - and also Bug 1182959 changes for webaudio on this tree (d15675a89) - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1181452 - Add memory reporting infrastructure for Servo to SpiderMonkey. r=till. (0c503da32) - Bug 1185739 - Fix self-comparison warnings in MemoryMetrics.h. r=efaust (08da5402c) (06bd03cd6) - ported changes from mozilla: - Bug 1182969 - Use nsTHashTable::Iterator in accessible/. r=tbsaunde. (9e8144b17) * 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