Leaderboard

And here is another important note! In order to install my new mod uBlock Origin Legacy 1.16.4.32 successfully and without any errors, please follow the instructions provided in my article! Here are the essential points once again: Backup all settings via the "Back up to file"" function in uBlock Origin. Completely uninstall the previous uBlock Origin version. Clear the browser cache. Close the browser. Manually delete the ublock0.sqlite file in your profile folder. Restart the browser. Install my mod. Import the settings from the backup file. Starting with the version uBlock Origin Legacy 1.16.4.32, my mod has a new ID. Therefore, a clean install is mandatory. Cheers, AstroSkipper

Version 1.16.4.32 works flawlessly - thanks again! Since uBO Legacy is important for many, many users I hope you will consider in the future uploading it somewhere where it can be easily installed and kept automatically up to date (e.g. Pale Moon Add-ons Site, GitHub or wherever you see fit). A suggestion for your consideration! There is a nice little improvement in the webextension, in the "Trusted sites" section. Default entries there have a different color and are separately sorted from the user-created entries. In the Legacy version, everything is thrown in together which can be a little confusing if many entries are present. Not an earth-shattering feature, but if it's easy to implement it'd be nice to have.

Like @NotHereToPlayGames always says, we'd like to see the documented measurements. No disrespect to win32, but I see the 121 version is much slower and buggy, comparing to the fast'n'good 117. Probably, it has something to do with the XP porting. And I'm not the only one noticing. https://msfn.org/board/topic/185045-supermium/?do=findComment&comment=1259143

Yes, you'd do better were to stop offending people, particularly due to the profanity in your second language, and keeping the high road. It does exist. https://msfn.org/board/topic/185045-supermium/?do=findComment&comment=1259172 I use Dixel's launcher, it prevents Supermium from storing crap in the LocalState file.

Major, important, fundamental flags like --no-first-run or --time-zone-for-testing=UTC, they didn't change, so no worries. What was a pleasant surprise, you study right from me, so maybe not all is lost. I took a quick look at your list, your research is far better than the same of @Sampei.Nihira.

you need to delete startupCache folder in your profile in order to activate the manual fix.

because there is a late commit fixing this: https://github.com/roytam1/UXP/commit/1116e8e137a3525b8313505a71894bb50af816d6 you may need to fix this by hand or wait for next build.

Well, it's the first build for XP, the alpha. Seeing the thoroughness of its development, I trust newer versions will do better. Bottom line: the browser's settings have to be reviewed to prevent such a heavy first load in XP, it should review the requirements/resources allowed to GG flagship apps, and also - not reported before- the persistence of chrome processes after closing it, even when the settings command it to do so. I forgot to mention: I ran it with umatrix, ublock (full set of lists, as in Astroskipper's legacy + a few more), clearurl, behinf the overlay, singlefile and tampermonkey). I will keep the browser as it stands with my tweaks now, because it may allow me to enter sites I can't with sp52, mypal68, or the last mod of 360chrome by Articfoxie.

SPEEDOMETER 2.0 SCORES W7 = Supermium 117 = 88 XP = Supermium 121 = 91 XP = chrome M115 = 63 XP = 360 chrome 13.5 = 45

But these fonts aren't any better than others reported.

I can't tell about Brave 121, but I compared Brave 110 vs this one, and Brave at least has no leaks, no crashes. But it's on Vista. I'm sorry, I didn't document comparison charts, but Brave was faster, and the UI in Brave is much better to my personal taste.

I think I was one of the first ones to discover it, I had hopes for this browser. The heavy leak was there some versions ago already, probably from the beginning. That said, I have enough RAM, but it won't just stop. https://msfn.org/board/topic/175262-last-versions-of-software-for-windows-vista-and-windows-server-2008/?do=findComment&comment=1258182

Found a small bug: When hovering over the Control Center icon, the Control Center tooltip appears, but after a moment it is overlapped by the Bluetooth Devices tooltip (or the network tooltip when bluetooth is tuned off).

@anton12 As already described in previous posts, you can easily fix this bug yourself manually, strictly following the commit https://github.com/roytam1/UXP/commit/1116e8e137a3525b8313505a71894bb50af816d6 @roytam1 kindly posted. Works great here.

x58 with 2.5GB available memory This is a bit misleading -- Supermium lags/freezes for up to several secs when launched, memory use spikes. Once running, the numbers seem representative. Mypal 68 benches similarly to Chrome backports. <-click See above. It's a memory hog, like most full-featured modern browsers.

So how it correlates with what you wrote before?

The list needs to be rechecked, from a short stare, I can see: some flags were added several times --process-per-site, for example.

You're right. It is not the install.rdf file. The string uBlock Origin Legacy inside this file is what is shown in the Add-ons Manager. Anyway! There are a couple of files which have to be modified, and it is more complicated as I originally thought. gorhill has created a very special, atypical and complex extension. Due to my experiences that very bad things can happen if modifying important or sensible files inside the uBlock Origin xpi file, I won't provide details about my way of modification. Please understand that! The consequence would be in the end that everyone would change something, and nothing would work afterwards. And then I have to listen to things here or elsewhere that weren't my fault. I'm not in the mood for that at all. In contrast, I definitely know what I'm doing. It's as simple as that. However, one thing is clear. Anyone can view the code of this extension themselves. It is open code and readable. Anyone with a little more in-depth knowledge knows how to identify the changes. And they are probably rather the right ones to change things. Cheers, AstroSkipper

In my revised and fixed version of uBlock Origin Legacy 1.16.4.32, the issue with the empty buttons should not happen anymore. If you have further problems with my mod, or you want to post about your experiences with my mod, I would be very happy if you did that here first. Thanks! Greetings from Germany, AstroSkipper

@billmcct The thread "New uBlock Origin version released (Special mod by @AstroSkipper)" in the Pale Moon forum is not the support thread of my mod uBlock Origin Legacy. By the way, this thread is badly moderated and contains a lot of off-topic comments. If you had been following the thread here closely, you would have noticed that I had found a bug which I had already fixed last night, tragically at considerable expense of time. So, as I have already announced here, you will have to download my latest version 1.16.4.32 again.

<OT> All of the goat talk reminded me of this -- https://www.youtube.com/watch?v=xrTUmYxnNlo </OT> re: "Tough crowd". Yep, MSFN seems to always be a "tough crowd". Good and bad.

If you were talking about Mypal 68, then this would be unfortunately off-topic here. This thread is about legacy extensions for UXP-based browsers. Don't forget we have a thread I created especially for extensions in Mypal 68!

Raising Hope 2010

Epic 120 developed in INDIA scored 64 https://epicbrowser.com/thank-you

I did almost the same thing and disabled just about every and any flag I thought was slowing it down or that which was not needed for my basic browsing needs but it made no difference sad to say - perhaps it will improve over time. The only switches I tried were: --no-first-run --disable-file-system --disable-gpu --disable-logging --disable-webg --disable-component-update --disable-background-networking --ssl-version-min=tls1.2 --enable-strict-mixed-content-checking --no-pings I figured --disable-file-system --disable-gpu should improve it, but no - keeping in mind my hardware is so old better to not try to enable acceleration that would be pointless. Interesting to test this new browser anyway regardless of the outcome.

I said I tested here and there, including MSFN (italics added for emphasis). That means MSFN was one of many tabs I worked with, while testing. Comprende? As per testing YT or GM first: folly. If you read back, you'll see it took a while for me to get supernium not to crash on start. When I achieved that, I started disabling things I know cause loading issues. When I manage to make it load, seeing it was still heavy, I disabled a bunch of chrome//flags. All the while reporting and surfing around, to see what happened, including hog FB. Once I couldn't think of more things to disable/turn off, only then I tried YT (in many ways) with no success (always crashing the tab or the browser, and then GM, which resulted unusable, for slow. And with that, my testing of this version of Supernium ended.

Thanks 66cats for screenshots.

https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20240127-3219d2d-uxp-35252384aa-xpmod-ia32.7z Hello VistaLover, You are right. The latest (s.a.) St52 shows the same bug-pattern as NM28. Just take my latest post on NM28 concerning this bug and replace St52 for NM28 - the same story.

Not sure what you mean. If we're talking about XP, it's the only version of Supermium & is substantially faster than other XP-compatible browsers on decent HW, both subjectively & benched by Speedometer 2.0: Mypal 68 and Supermium, side-by-side on x99 (x64) On x58 (x64) On Haswell (x86) I use Google & like having sync. Different set of priorities.

Agreed! I see the same thing here. Granted, I have *barely* tested Supermium. That goes for 117 or for 121 or anything in-between. It is really "not for me" until it becomes "ungoogled". What I myself would have preferred to see is a v115 (first version for Supermium) being updated, vetted by beta testers, updated again, vetted again, improved, vetted again, optimized, vetted again all while remaining at v115. For v115 to have made it to maturity (stable on XP, UNGOOGLED, et cetera). Then, only only then, for a v121 to be released with the same set of patches that brought v115 to maturity. Granted, I am a self-admitted biased Chromium User. I do NOT use the latest-and-greatest if it shows itself to be SLOWER than PREVIOUS versions. And YES, not only by "gut feelings" but by quantifiable, repeatable, "speed test" metrics. The two DO go hand-in-hand much more than the "I don't support those types of tests" folks care to bring themselves to admit. My favorite still remains v114 which predates "all" of the Supermium releases.

I learn from everybody who has something of value to offer. You have made your valuable contributions, but they are (in my opinion, no offense intended, just the facts) of no comparison to those of @Sampei.Nihira. You'd do better were to stop offending people, particularly due to their proficiency in a second or third (for some of us, fifth or more) language, and keeping the high road. It does exist. Back on topic: in yourprofilepath, next to the default folder, one can find a file called "Local State." It holds all the chrome://flags you move out of default. As I was saying, " I'm systematically disabling flags related to: autofill, webrtc, webgl (now unnecessary, chrome finally caught up), payment, credit (reviewing card with detail, for there are some related to cards other than credit cards and smart card), share/sharing, password (most of them), sync (most of them), geolocation, and history (most of them)." Those choices get saved in "Local State". If you save that file (together with your extensions and browser settings in the "Default" folder, you can then recreate your profile from scratch, with your preferred stuff and preferences, fresh as the same day you finished customizing your browser. Which I did today. Cheers!

True that. I also noticed I added a --site-per-process in there. It doesn't seem to hurt, but who knows. For my use it doesn't do damage. All suggestions/corrections are welcome. Supernium itself recommended dropping ones I was using in 360 chome. As per chrome://flags, many of the old ones that I used in 360chrome are not there anymore. I'm systematically disabling flags related to: autofill, webrtc, webgl (now unnecessary, chrome finally caught up), payment, credit (reviewing card with detail, for there are some related to cards other than credit cards and smart card), share/sharing, password (most of them), sync (most of them), geolocation, and history (most of them).

What is good, at least you learn from me the right variants (my post link), and not replicate the errors made in @Sampei.Nihira's list.

... FWIW, this very same bug can be observed in the latest St55 (32-bit, buildID=20240125063308) ; I suspect the "fix" is similar, at least according to: https://github.com/roytam1/basilisk55/commit/323bda6dcc13f3875ef36bbce15b59107ac0f09e BTW, I haven't updated my daily driver, St52, but I expect the bug to be there, too, since it affects ALL UXP-based apps ...

Dave, does the new mobo has the same BIOS?

Thanks! However, after a fresh install of New Moon 28, it worked without deleting the startupCache folder. But good to know. I usually delete this cache every time if I change CSS stylesheets in Mypal 68.

Thank you very much for clarification! I already thought that my New Moon installation was faulty.

Arguably more secure than an "insecure" web browser on an "insecure" OS.

Thanks for your report! I totally forgot to mention this issue. Same problem in my installation. Once opened and closed then, the Browser Console can't be opened again. So, I can confirm your observation. But it is very annoying that the Browser Console opens automatically every time the browser is started as soon as it is called up just once.

As per the loader, the relevant parts are [Setup] AppName=chrome [Environment] Profile=YOURPATHTOPROFILE [FileToRun] PathToExe=.\Supernium\$AppName$.exe Parameters=--user-data-dir="%Profile%" --no-first-run --no-default-browser-check --no-referrers --no-pings --disable-component-update --disable-logging --disable-background-networking --allow-outdated-plugins --disable-client-side-phishing-detection --safebrowsing-disable-download-protection --disable-background-mode --disable-quic --disable-suggestions-service --kiosk-printing --disable-print-preview --load-media-router-component-extension=0 --ssl-version-min=tls1.2 --enable-strict-mixed-content-checking --process-per-site --renderer-process-limit=3 --disable-breakpad --disable-machine-id --block-new-web-contents --disable-crash-reporter --disable-default-apps --disable-domain-reliability --disable-external-intent-requests --disable-file-system --gpu-rasterization-msaa-sample-count=0 --lite-video-force-override-decision --num-raster-threads=4 --process-per-site --site-per-process --time-zone-for-testing=UTC --use-fake-device-for-media-stream --enable-features="BlockInsecurePrivateNetworkRequests,BlockInsecurePrivateNetworkRequestsForNavigations,BrowserDynamicCodeDisabled,DesktopScreenshots,DisableProcessReuse,ElementSuperRareData,EnableCsrssLockdown,EncryptedClientHello,ForceIsolationInfoFrameOriginToTopLevelFrame,GpuAppContainer,ImprovedCookieControls,IntensiveWakeUpThrottling:grace_period_seconds/10,IsolateOrigins,IsolatePrerenders,IsolateSandboxedIframes,MinimizeAudioProcessingForUnusedOutput,NetworkServiceSandbox,NetworkServiceCodeIntegrity,OpaqueResponseBlockingV01,OriginIsolationHeader,PartitionConnectionsByNetworkIsolationKey,PartitionDomainReliabilityByNetworkIsolationKey,PartitionExpectCTStateByNetworkIsolationKey,PartitionHttpServerPropertiesByNetworkIsolationKey,PartitionNelAndReportingByNetworkIsolationKey,PartitionSSLSessionsByNetworkIsolationKey,PartitionedCookies,PostQuantumCECPQ2,PrefetchPrivacyChanges,ReduceUserAgent,ReducedReferrerGranularity,RendererAppContainer,RestrictGamepadAccess,SandboxExternalProtocolBlocked,ScopeMemoryCachePerContext,SplitAuthCacheByNetworkIsolationKey,SplitCacheByIncludeCredentials,SplitCacheByNetworkIsolationKey,SplitHostCacheByNetworkIsolationKey,StrictOriginIsolation,SubframeShutdownDelay,SuppressDifferentOriginSubframeJSDialogs,ThirdPartyStoragePartitioning,ThrottleForegroundTimers,TurnOffStreamingMediaCachingAlways,TurnOffStreamingMediaCachingOnBattery,WinSboxDisableExtensionPoint,WinSboxDisableKtmComponent" --disable-features="AcceptCHFrame,AdInterestGroupAPI,AllowClientHintsToThirdParty,AllowURNsInIframes,AnonymousIframeOriginTrial,AutofillEnableAccountWalletStorage,AutofillServerCommunication,BrowsingTopics,ClearCrossSiteCrossBrowsingContextGroupWindowName,ClientHintThirdPartyDelegation,ClientHintsDPR,ClientHintsDPR_DEPRECATED,ClientHintsDeviceMemory,ClientHintsDeviceMemory_DEPRECATED,ClientHintsMetaHTTPEquivAcceptCH,ClientHintsMetaNameAcceptCH,ClientHintsResourceWidth,ClientHintsResourceWidth_DEPRECATED,ClientHintsViewportWidth,ClientHintsViewportWidth_DEPRECATED,ComputePressure,ContextMenuPerformanceInfoAndRemoteHintFetching,ConversionMeasurement,CookieDomainRejectNonASCII,CopyLinkToText,CrashReporting,CriticalClientHint,CrostiniAdditionalEnterpriseReporting,CssSelectorFragmentAnchor,DocumentReporting,EnableStructuredMetrics,EnterpriseRealtimeExtensionRequest,ExpectCTReporting,EnableTLS13EarlyData,FedCm,Fledge,FontAccess,GreaseUACH,IdleDetection,InterestGroupStorage,Journeys,LensStandalone,MediaDrmPreprovisioning,MediaEngagementBypassAutoplayPolicies,NavigationRequestPreconnect,NetworkTimeServiceQuerying,NotificationTriggers,OmniboxTriggerForNoStatePrefetch,OptimizationHints,OptimizationHintsFetching,OptimizationHintsFetchingAnonymousDataConsent,OptimizationHintsFieldTrials,Parakeet,Prerender2,PrefersColorSchemeClientHintHeader,PreloadMediaEngagementData,Reporting,RetailCoupons,SCTAuditing,SegmentationPlatform,SignedExchangeReportingForDistributors,SignedHTTPExchange,SpeculationRulesPrefetchProxy,SubresourceWebBundles,TabMetricsLogging,TFLiteLanguageDetectionEnabled,TextFragmentAnchor,SafeBrowsingBetterTelemetryAcrossReports,UserAgentClientHint,UserAgentClientHintFullVersionList,UsernameFirstFlow,UsernameFirstFlowFilling,UsernameFirstFlowFallbackCrowdsourcing,ViewportHeightClientHintHeader,WebNFC,WebOTP,WebSQLInThirdPartyContextEnabled,WebXR,WinrtGeolocationImplementation,RecordWebAudioEngagement" --cipher-suite-blacklist=",0x000A, 0xc013,0xc014,0xe013,0x009c,0x009d,0x002f,0x0035" --component-updater=url-source=0.0.0.0 --crash-server-url=0.0.0.0 --gcm-checkin-url=0.0.0.0 --gcm-mcs-endpoint=0.0.0.0 --gcm-registration-url=0.0.0.0 --google-apis-url=0.0.0.0 --lso-url=0.0.0.0 --oauth-account-manager-url=0.0.0.0 --override-metrics-upload-url=0.0.0.0 --realtime-reporting-url=0.0.0.0 --reporting-connector-url=0.0.0.0 --sync-url=0.0.0.0 --variations-server-url=0.0.0.0 --variations-insecure-server-url=0.0.0.0 --cipher-suite-blacklist="0xc013,0xc014,0x009c,0x009d,0x002f,0x0035" --blink-settings="dnsPrefetchingEnabled=false,preferredColorScheme=1,strictMixedContentChecking=true,strictMixedContentCheckingForPlugin=true,strictlyBlockBlockableMixedContent=true" --isolation-by-default --js-flags=--jitless --disable-webgl --component-updater=url-source=0.0.0.0 --enable-features="RendererAppContainer"

You claim peasant slang to be imperial tongue , well, go on like this, and I might follow this topic, boy. Hardware acceleration of 10bit HEVC {H265} is not available throughout the whole 700 series, except 750SE, a super budget, rare and outdated card. It's been explained here. As for 8bit HEVC, no one uses it anymore for encoding, no one, so nothing to play. I actually tried an old film encoded in H265 Main 8, it's choppy with this player.

Indeed! WhoTracks.me's script puts a heavy load on my Core Quad computer bought for 2000€ in 2007, surely looks like someone who gave the advice to use WhoTracks.me has a modern PC with Windows 10-11. I guess 4000€ now is the same as 2000€ in 2007.

New build of Firfox 45ESR SSE: test binary: https://o.rthost.cf/gpc/files1.rt/firefox-45.9.11-20180916-ca29a3ea9-win32-sse.7z repo: https://github.com/roytam1/mozilla45esr Changes since my last build: - import changes from tenfourfox: - #516: make timeouts less aggressive (4c83f3281) (M1352874 is already ported in rev 59c46aa86) - #512: M1469348 M1478575 M1461706 (c27f5bf59) - #512: update certs and pins (9b1dfff49) (36bc50793) - update NSS to rev 8f6014565b91 with nss339-vc2013.diff applied (3c2526292) - bump revision to 45.9.11 (f1d0cbb7c) - import changes from tenfourfox: - #399: implement Event.srcElement as alias (2c14dc434) - closes #519: Element.toggleAttribute() + backbugs M1258205 M1276938 (e5c93de89) - closes #520: Document.elementsFromPoint() M1164427 (7e7cdb504) (ca29a3ea9) Known issue: sites providing TLS 1.3-draft18 may not work properly. To workaround this, change security.tls.version.max to 3 in about:config.

New build of basilisk/UXP for XP! Test binary: Win32 https://o.rthost.cf/basilisk/basilisk52-g4.1.win32-git-20180915-351ffa462-xpmod.7z Win64 https://o.rthost.cf/basilisk/basilisk52-g4.1.win64-git-20180915-351ffa462-xpmod.7z diff: https://o.rthost.cf/basilisk/UXP-xp-gitdiff-20180606.7z PM28XP build: Win32 https://o.rthost.cf/palemoon/palemoon-28.1.0a1.win32-git-20180915-351ffa462-xpmod.7z Win64 https://o.rthost.cf/palemoon/palemoon-28.1.0a1.win64-git-20180915-351ffa462-xpmod.7z Official repo changes since my last build: - Make nsScannerIterator a trivial type in order to use memset (d79c47502) - Convert the trailing array of BindingNames at the end of the various kinds of scope data into raw unsigned chars into which those BindingNames are placement-new'd, rather than memcpy-ing non-trivial classes around and failing to comply with the C++ object model (b4aed63f5) - Call the relevant scope-data constructor when allocating it, and poison/mark as undefined the memory for the trailing array of BindingNames, ratther than impermissibly PodZero-ing non-trivial classes. (36cb80d1c) - Fix build bustage (47c5bba17) - backport mozbug 1350090: Turn off the spammy warning that goes off every time we create an about:blank content viewer; r=mystor (22bb99a32) - RTCCertificate.cpp: output truncated before the last format character (cb7141536) - Merge pull request #757 from trav90/buildsystem-work (347e6ac31) - Merge pull request #755 from g4jc/debugspam (a90b6fe53) - Silence the -Wuninitialized warning in mozjemalloc (41b674505) - Merge pull request #758 from trav90/buildsystem-work (eb2016063) - Revert "Add tabbrowser findbar styling for Linux and OSX" (9900def52) - Revert "Add tabbrowser findbar styling" (d9f762df8) - Revert "[PALEMOON] Findbar - added the listener also to tabbrowser.xml, remove "_fastFind", a better focus() for the method "updateCurrentBrowser"" (95835c0ac) - Revert "[PALEMOON] Findbar - remove unnecessary code (Findbar is no longer in the gBrowser)" (b1a98da75) - Revert "Transfer findbar data to torn-off tabs." (aa9cd0d77) - Revert "Minimal easy fix -- move findbar getters to tabbrowser." (1c455d9cb) - Ensure the findbar target content browser follows the active tab. (a744d7665) - Add findbar.termPerTab (7c9b62ee0) - fix mozilla regression in search service (saving user-defined search engines) (2e9c525a9) - Bug 1470260 - Part 1: Ensure that 'this' stays alive for the duration of the TickRefreshDriver call. (4c1f33b16) - Bug 1470260 - Part 2: Make RefreshDriverTimer ref-counted and hold a strong ref on it on the stack when nsRefreshDriver::Tick can be reached. (69b88dfcc) - Merge pull request #754 from trav90/class-memaccess-errors (7bf3d2440) - Merge branch 'master' into ported-upstream (756551355) - Cherry-pick libvpx upstream 52add5896661d186dec284ed646a4b33b607d2c7. (ca36e991e) - Bug 1473113 - Defer initializing the MAR index until it's needed. (ea8680907) - Replace the custom logic in ObserverList with an nsTObserverArray which has all the necessary logic for stable iteration over a potentially changing list of items. (e05705fe5) - Remove some unnecessary dependencies from HAL. (1bf4a513c) - Bug 1478575 - Unify CamerasChild shutdown paths. (1de1ce577) - Bug 1467363 - Protect access to mTransparentSurface with a lock. (6ded94d38) - Move surface data checking to a separate function to make it less "totally nuts" (cbfef7fcd) - Release mapped surface memory on size check failure to plug a leak. (09319b435) - Fix build bustage in cbfef7fcdb853916ff04015f6ee2d4b86f424a08 (6195a2f1e) - Bug 1469309 - Remove an unused sensor type. (a95078a6c) - Bug 1472925 - Keep a strong reference to MediaStreamGraph from GraphDriver. (9830cd079) - Bug 1466991 - Part 1: Factor out ShouldUseXBLScope. (a6d4e65db) - Bug 1466991 - Part 2: Reparent nodes when they start being in the XBL scope. (d086e918b) - Merge branch 'ported-upstream' (2713347f0) - Merge branch 'fbgw' (b07b08f9d) - Update HSTS preload list (2f64d5eee) - Flush some more buildlog output to screen when prudent. (9f091cf11) - Fix clang build bustage. (2bbbcfc49) - Merge branch 'master' of https://github.com/MoonchildProductions/UXP (2fce053e9) - Fix wrong SVG sizes with non-integer values for viewBox width/height. (351ffa462)

New build of basilisk/UXP for XP! Test binary: Win32 https://o.rthost.cf/basilisk/basilisk52-g4.1.win32-git-20180908-b728af022-xpmod.7z Win64 https://o.rthost.cf/basilisk/basilisk52-g4.1.win64-git-20180908-b728af022-xpmod.7z diff: https://o.rthost.cf/basilisk/UXP-xp-gitdiff-20180606.7z PM28XP build: Win32 https://o.rthost.cf/palemoon/palemoon-28.1.0a1.win32-git-20180908-b728af022-xpmod.7z Win64 https://o.rthost.cf/palemoon/palemoon-28.1.0a1.win64-git-20180908-b728af022-xpmod.7z Official repo changes since my last build: - Revert "Switch to using a single memory allocation arena" (69627ad41) - Remove support for TLS session caches in TLSServerSocket. (1425f020c) - [PALEMOON] Update the taskbar preview per tab module for UXP. (35f68645a) - Remove FxA migrator. (2781afdf9) - [ffvpx] Update readme (45f9a0daa) - Remove all C++ Telemetry Accumulation calls. (ab961aeb5) - Merge pull request #745 from MoonchildProductions/Kill-Telemetry (2e00eb87e) - Update 'browser.newtab.url' when it's used for new tabs and 'browser.startup.homepage' changes (5359706e7) - Use the default value from 'browser.startup.homepage' instead of the hardcoded 'http://start.palemoon.org/' (fa764a68d) - Add extra check for assembler buffer space. (7d73b3fbf) - Merge branch 'master' of https://github.com/MoonchildProductions/UXP (45ec2bceb) - Remove all C++ telemetry autotimers (93cae908b) - Remove all C++ telemetry autotimers (b28ab55f9) - Part 1. Move IsLocalRefURL to nsContentUtils to reuse this function. (port-rewrite) (19f010c62) - Revert "Part 1. Move IsLocalRefURL to nsContentUtils to reuse this function. (port-rewrite)" (b43a1afcd) - Part 1. Duplicate IsLocalRefURL to nsContentUtils to reuse this function. (32464a0b9) - Part 2. Implement nsSVGEffects::GetBaseURLForLocalRef to export local-ref-url-resolving logic. (d314fcbcc) - Part 3. Resolve local-ref in SVGUseElement::LookupHref by nsSVGEffects::GetBaseURLForLocalRef. (246edb533) - Part 4. Reftest for using local-ref as xlink:href value. (6a7071fde) - Bug 1357432 (#750) (ae432c636) - Write out serialized search engine files when rebuilding cache. (bbdd1c931) - Merge branch 'Bug_1357432' (ddaf09fa2) - Merge pull request #747 from JustOff/PR_NewtabUrl (6c4be6259) - Bump goanna version for SVG handling update (871530bfe) - Merge branch 'master' of https://github.com/MoonchildProductions/UXP (d31ab5ff4) - [GCC] Add -Wno-noexcept-type to the js build flags (346ba48dd) - Merge pull request #753 from trav90/buildsystem-work (b728af022)

New build of basilisk/UXP for XP! Test binary: Win32 https://o.rthost.cf/basilisk/basilisk52-g4.1.win32-git-20180901-4f504d46b-xpmod.7z Win64 https://o.rthost.cf/basilisk/basilisk52-g4.1.win64-git-20180901-4f504d46b-xpmod.7z diff: https://o.rthost.cf/basilisk/UXP-xp-gitdiff-20180606.7z PM28XP build: Win32 https://o.rthost.cf/palemoon/palemoon-28.1.0a1.win32-git-20180901-4f504d46b-xpmod.7z Win64 https://o.rthost.cf/palemoon/palemoon-28.1.0a1.win64-git-20180901-4f504d46b-xpmod.7z Official repo changes since my last build: - Refresh nsStringBundleService and nsHttpHandler when the browser locale is changed (02211f844) - Add a horizontal scroll action option for mouse wheel. (f0e053a1b) - Switch to using a single memory allocation arena (4ceb21241) - Add missing } (4cbcba787) - Merge pull request #733 from JustOff/PR_locale_changed_refresh (2dc1703c0) - Don't build ffvpx on 32-bit platforms (except Windows) (a64465bde) - Update installer strings for Pale Moon to match toolkit nsis installer. (f6d3d16ac) - Update HSTS preload list (cc0e6c7ec) - Sync disabled state of number control regardless of appearance. (693656e99) - Fix incorrect code removal in ipc. (0bc251143) - Merge pull request #741 from trav90/media-work (f266d26eb) - Revert "Bug 1444668 - Avoid allocating large AssemblerBuffers. r=luke, r=bbouvier, a=RyanVM" (f214aa5db) - Merge branch 'master' of https://github.com/MoonchildProductions/UXP (1edc7328f) - Revert "Bug 1388020. r=nical, a=RyanVM" (4f504d46b)

PM27 is faster than PM28 about browsing facebook timeline, but still strugglish when comparing with blink-based browsers.

New build of post-deprecated basilisk/moebius for XP! * Notice: This repo will not be built on regular schedule, and changes are experimental as usual. ** Current moebius patch level should be on par with 52.8, but some security patches can not be applied/ported due to source milestone differences between versions. Test binary: Win32 http://o.rths.cf/basilisk/basilisk55-win32-git-20180828-b78c31157-xpmod.7z Win64 http://o.rths.cf/basilisk/basilisk55-win64-git-20180828-b78c31157-xpmod.7z repo: https://github.com/roytam1/basilisk55 Repo changes: - cherry-picked from UXP: Improve nsHtml5AtomTable performance (#693) (dc5f43707) - update NSPR to hg rev 6e31156d7002 (3c015fea6) - update NSS to hg rev 27e7c0376137 (f791dfd70) - update ffvpx to git release/3.4 branch rev bc2301429e9c779237e43acf913331af018211f2 (12091e9f0) - update NSS to hg rev 4c7ffcfd43f6, with TLS 1.3 RFC version enabled. (b78c31157) * This is a special build intended to test NSS 3.39 Beta about TLS 1.3 RFC version

New build of Firfox 45ESR SSE: test binary: https://o.rthost.cf/gpc/files1.rt/firefox-45.9.10-20180828-169242048-win32-sse.7z repo: https://github.com/roytam1/mozilla45esr Changes since my last build: - update NSPR to hg rev 34274ae8c85e (ed341512b) - update NSS to hg rev 4c7ffcfd43f6 (320507fa1) - apply VC2013 fix for NSS 3.39 (169242048) * This is a special build intended to test NSS 3.39 Beta about TLS 1.3 RFC version

New build of Firfox 45ESR SSE: test binary: https://o.rthost.cf/gpc/files1.rt/firefox-45.9.10-20180825-fc7578166-win32-sse.7z repo: https://github.com/roytam1/mozilla45esr Changes since my last build: - import changes from tenfourfox: - more hosts for adblock (0558c0179) - #517: fake out current Rocket Loader (0bbf06ea6) - #517: a couple tune-ups (5f47da230) - #512: our own fix for M1412081 (c469554e9) - #512: M1480092 M1466577 (812d03b96) - #518: eat our shorts, github (43534dea3) (fc7578166)