Leaderboard

I respect your opinion, brother. Naturally, people's use cases will vary. FX, PM, Basilisk, etc., like any other browser, will never be 'perfect'. For me, I've had very good results over the years with most of the builds I've tried, but a lot of that can be attributed to keeping JavaScript turned off as much as possible, which seems to make a big difference towards speed and security. Of course, Gecko/Goanna/UXP have had problems in recent years with bloated code bases. Especially if you keep JavaScript turned on, web browsing tends to be a slow experience as the aging JS engines are running into more challenges with badly designed web pages and their abundance of badly coded HTML5, but for all their flaws Gecko/Goanna/UXP still do a respectable job with the modern web. If I could use something like Firefox 3.6, RetroZilla or Netscape 9 on a daily basis as my all-purpose browser then I absolutely would, because it is frustrating to see how web browsers have had to adjust to accommodate all of these bells/whistles that the incompetent web designers ('script-kiddies') of today like to add to their pages. There are literally no advantages to how most sites are designed today versus what was possible for browsers to handle back in 2008. (I could go on, but I don't want to rant.) Even with its flaws, I'm still using MyPal 29.3 as my main browser, but MyPal 68 is very useful as a secondary option, and I'm happy that both versions can coexist with no problems in the same XP installation. Unless MyPal 29.3 becomes too outmoded and dated for daily use, it will likely remain my primary browser for the foreseeable future. It will be exciting to see what improvements are made to MyPal 68 going forward, and what will happen with future MyPal versions.

New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20220521-3219d2d-uxp-1e871780f-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20220521-3219d2d-uxp-1e871780f-xpmod.7z source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/custom IA32 Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20220521-3219d2d-uxp-1e871780f-xpmod-ia32.7z source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/ia32 NM28XP build: Win32 https://o.rthost.win/palemoon/palemoon-28.10.6a1.win32-git-20220521-d849524bd-uxp-1e871780f-xpmod.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.6a1.win32-git-20220521-d849524bd-uxp-1e871780f-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.6a1.win64-git-20220521-d849524bd-uxp-1e871780f-xpmod.7z Official UXP changes picked since my last build: - Issue #1899 - Disable the (broken) MDN integration widget by default. (0749a8828) - Issue #1899 - Make sure the test for it still works (349d3d96a) - Issue #1898 - Make sure that the sanity test stops running if necessary (d42af42d2) - Issue #1813 - Enable date and time picker by default. (40e9f40e8) - Issue #1210 - Keep timepicker disabled for now. (8dfcd4553) No official Pale-Moon changes picked since my last build. No official Basilisk changes picked since my last build. My changes picked since my last build: - GMP: revert GMPUtils.jsm back to state of rev 4a010b94951a46c3f85da43b1fd712fb1a8f8bc5 (ed4e67920) * Notice: From now on, UXP rev will point to `custom` branch of my UXP repo instead of MCP UXP repo, while "official UXP changes" shows only `tracking` branch changes.

I've prepared test build with restyled dark acrylic menus, please test http://startisback.com/StartAllBack_setup.exe

I suspect whatever flaws are lurking there wouldn't be as pronounced if web sites were designed differently. Many sites are unusable without JavaScript. I wonder what would be a good content rich website that works without JavaScript and if leaks are really mostly due to JavaScript. Since these browsers do have a loyal following of developers, surely someone must have thought about checking with debugger where memory allocations stay? Everything develops at really fast rate, much longer development cycles would be a whole lot better in the general. But we live in this insane hyper-capitalist society that keeps chasing its own tail and can't slow down for one minute and take a good look around to see where the heck it's at.

And one additional tip. For better overview delete all unnecessary entries in config.ini! Keep only entries of sites you really use! @heinoganda added a lot of sites, maybe to his needs. The config.ini of HTTPSProxy is nearly empty comparing to ProxHTTPSProxy's one. Cheers ,AstroSkipper

Problem solved. Config.ini has to be modified. In section [SSL Pass-Thru] you have to deactivate *gov.uk* by setting # in front of the string or delete *gov.uk* completely.

I've read that JustOff is Ukrainian, and so would have bigger issues than maintaining browser add-ons to deal with these days. One hopes he is well and safe....

Agreed! That really and truly is the only end possible with an ecosystem that revolves around "nightly" and "weekly" updates. You cannot properly and adequately "test" software with that release schedule. Your user-base does the "testing" for you. Oten resulting in bugs from six months ago being buried very deep because it took six months for the user-base to find it.

yep, got 2018 off my memory.

Throughout the years, speaking from my own experience (FX->PM->FX->Basilisk->PM), it often felt like the codebase is held together by duct tape, eg. memory leaks continue to persist. You close all tabs, but the browser process still holds hundreds of megabytes of RAM. Or take Scratchpad editor's "Pretty print" function for instance, broken that it often breaks the code or craps itself; doesn't understand "let" variable declarations (very common!) and "`", there could be more valid characters / words. This doesn't make sense, ANGLE enables D3D9/D3D11 output (only former usable on XP), with "webgl.disable-angle" set to true, OpenGL is used, not D3D9 or D3D11.

News: nullish coalescing operator PR is created by FranklinDM in upstream, hopefully available in next build: https://repo.palemoon.org/MoonchildProductions/UXP/pulls/1901

Fine by me, but I'll bring this to the attention of @Tripredacus and @dencorso in case they have any reservations about it.

New NewMoon 27 Build! 32bit https://o.rthost.win/palemoon/palemoon-27.10.0.win32-git-20220521-e7535dc94b-xpmod.7z 32bit SSE https://o.rthost.win/palemoon/palemoon-27.10.0.win32-git-20220521-e7535dc94b-xpmod-sse.7z 32bit noSSE https://o.rthost.win/palemoon/palemoon-27.10.0.win32-git-20220521-e7535dc94b-xpmod-ia32.7z 64bit https://o.rthost.win/palemoon/palemoon-27.10.0.win64-git-20220521-e7535dc94b-xpmod.7z source repo: https://github.com/roytam1/palemoon27 repo changes since my last build: - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1194059 (Part 1) - Ensure that metadata decode progress is always delivered atomically. r=tn (07f0441600) - Bug 1191090 - Use the normal PNG decoder for PNG metadata decodes. r=tn (ce3fe1be5f) - Bug 1191114 (Part 1) - Always detect HAS_TRANSPARENCY during the metadata decode. r=tn (3841132932) - Bug 1191114 (Part 2) - Add support for creating an anonymous metadata decoder, for use in tests. r=tn (2cdcc0c278) - Bug 1191114 (Part 3) - Add flags to image test cases. r=tn (4a6f5a5230) - Bug 1191114 (Part 4) - Add tests for metadata decoding, including that we always deliver HAS_TRANSPARENCY during the metadata decode. r=tn (b9c5d1cd4a) - Bug 1126330 - Remove the check for non-looping animations. r=seth (828dabba24) - Bug 1194059 (Part 2) - Always detect IS_ANIMATED during the metadatadecode. r=tn (0ba5bf38f1) - Bug 1194059 (Part 3) - Ensure the nsIInputStream LoadImage() returns is always buffered. r=tn (ed2b02205b) - Bug 1194059 (Part 4) - Add tests that we detect IS_ANIMATED during the metadata decode. r=tn (298f14a7c9) - Bug 1188705 (part 1) - Remove gfxASurface::GetMemoryLocation(). r=seth. (1f0da73a08) - Bug 1188705 (part 2) - Remove unused SizeOfDecodedWithComputedFallbackIfHeap declaration. r=seth. (3356dbed06) - Bug 1188705 (part 3) - Simplify imgFrame::SizeOfExcludingThis(). r=seth. (563262a834) - Bug 1155252 - Don't allocate X11TextureClients bigger than xlib's maximum surface size. r=jrmuizel (3f11590667) - Bug 1143994 - Fix some -Wunreachable-code and -Wswitch warnings in imagelib. r=seth (008becc7e2) - Bug 1060609 (Part 1) - Disable downscale-during-decode when HQ scaling is disabled. r=tn (6da77e3cad) - Bug 1187569 - PNGs getting stuck in a pixelated state. r=seth (da305ef99c) - Bug 1194900 - Stop deciding when to send invalidations in nsPNGDecoder and let Decoder handle it. r=tn (50fa14a984) - Bug 1151694 - Part 1 - Move CommonAnimationManager::sLayerAnimationInfo into LayerAnimationInfo.(cpp|h). r=bbirtles (9f93e0d569) - Bug 1151694 - Part 2 - imgTools should be inside mozilla::image namespace. r=bbirtles (8dfc3f2e4b) - Bug 1196066 (Part 1) - Fix bad directory entries in two of our ICO reftests. r=tn (9e4c70d2b4) - Bug 1196065 - Add sanity tests for image decoders. r=tn (557b9131cb) - Bug 1194912 (Part 1) - Add CopyOnWrite<T> to support automatic copy-on-write for recursive writes to data structures. r=tn (b081a50716) - Bug 1196066 (Part 2) - Add a streaming lexing framework to ImageLib. r=tn (59eb634ea5) - Bug 1196476 - Replace ProgressTracker::FirstObserverIs() with a simpler mechanism on imgRequest. r=tn (db9ecc65ef) - missing part of Bug 1139225 (Part 2) - Dispatch OnImageAvailable to the main thread manually in imgRequest. r=tn (e7b22db614) - Bug 1194912 (Part 2) - Store ProgressTracker observers in a copy-on-write hash table, and dispatch notifications to them using a template. r=tn (5efd7b38b3) - Bug 1180225. Make convolver more like upstream. r=seth (18e3c168fc) - Bug 1149318 - Fix the calling convention on SkGetUserDefaultLocaleNameProc. r=eihrul (7b750d4e4e) - Bug 1210493 - enlarge stroke bounds by line width when doing a quick-reject in SkDraw::drawRect. r=jmuizelaar (e8b5d0fe2d) - Bug 1188206 - Fix more constructors in gfx; r=jrmuizel (944ea9938c) (b75b514a74) - import changes from `dev' branch of rmottola/Arctic-Fox: - namespace (dd7e1c593d) - Bug 1158541 - Remove TiledTextureImage::mSize; r=jrmuizel This member is never initialized to anything so it always contains a (0,0) size, and it shadows the protected mSize member in the base class which is actually initialized to the correct size. (e2930b05a8) - Bug 1158542 - Remove TextureImage::mImageFormat; r=jrmuizel This is never initialized or accessed. (c5af3763b5) - Bug 1136428 - Create a different set and restore path of the draw buffer state in WebGLContext::ForceClearFramebufferWithDefaultValues depending on whether or not the default framebuffer is being used. r=jgilbert (e92fbcaa38) - Bug 1136428 - Change implementation of WebGLContext::DrawBuffers such that it goes through GLScreenBuffer to change the draw buffer. Added a variable to GLScreenBuffer to cache the draw buffer mode. r=jgilbert (eb80c3ac18) - Bug 1136428 - Ensure we never create more than the amount supported of color attachments in WebGLFrameBuffer or try to get an attachment point higher than the supported number. r=jgilbert (3050ce6979) - Bug 1185803 - Replace MOZ_CRASH with GenerateWarning. r=jgilbert (1b5ba983c6) - Bug 1188540 - Forward texture-related functions to WebGLTexture. - r=kamidphish (10f84c83e5) - Bug 1189903 - Don't use RGBA surfaces on GLX if surface sharing is not used. r=jgilbert (2edbca3d13) - Bug 1170842 - Part 3: Implement GetInternalformatParameter. r=jgilbert, r=smaug (18f0fdf5f9) - Bug 1170842 - Part 4: Implement FramebufferTextureLayer. r=jgilbert (2acd7963e7) - Bug 1062066 (Part 1) - Add support for vertically flipping downscaler output. r=tn (f23315cba2) - Bug 1204626 - Add a regression test. r=billm (aa3c694b83) - Bug 1062066 (Part 2) - Add a Downscaler API to clear a row. r=tn (7c77bf27b0) - Bug 1062066 (Part 3) - Only mark BMP surfaces as transparent if they actually have alpha data. r=tn (ed9bade04e) - Bug 1060609 (Part 2) - Add downscale-during-decode support for the PNG decoder. r=tn,f=glennrp (47259a797a) - hack to fix build (12266daf48) - Bug 1060609 (Part 3) - Add tests that interlaced and non-interlaced PNGs have the same downscaling behavior. r=tn (763b18b15e) - Bug 1062066 (Part 4) - Add downscale-during-decode support for the BMP decoder. r=tn (5a810ed168) - indentation fixes (872c66d3a8) (00b0a024a4) - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1188569: Drop unneeded MOZ_WARN_UNUSED_RESULT from from LookupBestMatch in SurfaceCache.cpp. r=seth (5e74e0028c) - Bug 1192356 (Part 1) - Take advantage of mozilla::Tie() in SurfaceCache.cpp. r=dholbert (e4908c725d) - Bug 1192356 (Part 2) - Take advantage of mozilla::Tie() in RasterImage.cpp. r=tn (1204189b73) - Bug 1185800 - Add DecoderFlags and SurfaceFlags enum classes and use them instead of imgIContainer flags in all decoder-related code. r=tn (3abdab11f6) - Bug 1196066 (Part 3) - Rewrite nsICODecoder to use StreamingLexer. r=tn (e2ba590c9d) - Bug 1196066 (Part 4) - Enable the ICOMultiChunk test, which now passes. r=tn (9e02611959) - Bug 1124084 - Flip on downscale-during-decode everywhere. r=tn (bd9deff966) - Bug 1160801 - Treat invalid GIF disposal methods as DisposalMethod::NOT_SPECIFIED. r=jrmuizel (e26feaf8fb) - Bug 1201796 (Part 1) - Treat ICOs with wrong widths and heights as corrupt. r=tn (322ba20808) - Bug 1201796 (Part 2) - Add GetFrameAtSize() to support downscale-during-decode for GetFrame() use cases. r=tn (92f5d3a0a7) - Bug 1194906 - Replace 'NS_ENSURE_TRUE(BadImage(..))' warnings with more useful messages. r=tn (cc3b368673) - Bug 1201796 (Part 3) - Enable downscale-during-decode for imgITools::EncodeScaledImage(). r=tn (e2cdb5b520) - Bug 1194472 - Correctly fetch compositor backend in WebGLContext. r=jgilbert (0092052dfc) - Bug 1161913 - Part 1 - Add invalidation state for CaptureStream to Canvas and Contexts. r=mt (0377d6bbe7) - Bug 1168075 - Fix CanvasCaptureMediaStream build fail for bluetooth2. r=pehrsons (53c67c0056) - Bug 1176363 - Part 1: Make a raw copy of each Canvas::CaptureStream frame. r=mattwoodrow (a5df5793d6) - Bug 1194575 - Rename RecoverFromLossOfFrames() to RecoverFromInvalidFrames() to better reflect its role. r=tn (baa6455e79) - Bug 1146663 (Part 1) - Remove HQ scaling, which is now dead code. r=tn (efaddadea0) - Bug 1146663 (Part 2) - Remove the concept of lifetimes from the SurfaceCache. r=dholbert (ab9862d7ee) - Bug 1146663 (Part 3) - Make it impossible to deoptimize imgFrames. r=tn (19e2f1b370) - Bug 1201763 - Add downscale-during-decode support for the ICON decoder. r=tn (33a2b95e5c) - Bug 1194058 (Part 1) - Add Deinterlacer to allow Downscaler to work with interlaced images. r=tn (f7c57b7a8e) - Bug 1194058 (Part 2) - Add downscale-during-decode support for the GIF decoder. r=tn (85622f9d55) - Bug 1201796 (Part 4) - Add downscale-during-decode support for the ICO decoder. r=tn (d09d18b0d9) - Bug 1146663 (Part 4) - Make all RasterImages support downscale-during-decode. r=tn (264642a895) - Bug 1146663 (Part 5) - Require that all image decoders support downscale-during-decode. r=tn (79ad99885d) - Bug 1206836 - When downscaling ICOs, downscale the AND mask as well. r=tn a=KWierso (08ec3d092b) - missing bit of Bug 1138293 - Use malloc/free/realloc/calloc (eb8e5e1b9c) - missing bit of Bug 1146663 (Part 3) - Make it impossible to deoptimize imgFrames. (233befe48f) - Bug 1208935 - Move Deinterlacer to a standalone file. r=seth (b50322abc286) (f1d1e16669) - import changes from `dev' branch of rmottola/Arctic-Fox: - Bug 1187401 (Part 1) - Simplify the condition that determines whether we set RasterImage::mHasBeenDecoded. r=tn (69be36e7ef) - Bug 1187401 (Part 2) - Eliminate the nsresult return value from RasterImage::SetMetadata, since it's not used anymore. r=tn (dc521c4b9f) - Bug 1187401 (Part 3) - For consistency, call DoError if SetMetadata sees a negative size. r=tn (d9ca8ec91b) - Bug 1207183 - micro-optimize removing work items from DecodePool's queues; r=seth (6bd2717e3a) - No bug - Fix out-of-date comment in Decoder.cpp. r=me (54fdbbd444) - Bug 1181324 - Eliminate the duplicate mRefCnt member in MultipartImage. r=seth (c4512a443e) - Bug 1180105 - Do not leak the SourceSurface returned from imgIContainer::GetFrame in BlockUntilDecodedAndFinishObserving; r=seth (1cadbffc53) - Bug 1181909 - Fix potential null dereference in NextPartObserver. r=tn (cfd8ad0119) - Bug 108603 - Remove NS_IMPL_QUERY_INTERFACE_INHERITED0. r=mccr8, r=froydnj (4bfa6771fc) - Bug 1159502 - Don't block onload for multipart images. r=tn (8b50eadf39) - Bug 1200413 - Part 1: Re-write RefCountedInsideLambdaChecker to use captures instead of checking for DeclRef instances, r=ehsan (80ef99efe2) - Bug 1200413 - Part 2: Make lambdas in ProgressTracker.cpp capture strong references, r=seth (9e4d96dffa) - Bug 1194557 - Ensure that if the image was locked before RecoverFromLossOfFrames() was called, it's still locked afterwards. r=tn (ea4dc6ea9f) - code style (ad3773ba42) - Bug 1167590 - Mark imgRequestProxy::mListener as MOZ_UNSAFE_REF. r=seth (946ffaed8a) - Bug 1148397 - Fix data race on imgRequest::mHadInsecureRedirect. r=tanvi (e73d0664f3) - No bug - Tweak formatting of logging statement in imgRequest. r=me (cab2bcb014) - Bug 1180126 - Read content disposition regardless of content type in imgRequest::PrepareForNewPart. r=tn (2934597743) - Bug 1139225 - Followup - Remove duplicate multiPartChannel variable. (7f7f555a0b) - Bug 1141398 - Do not always revalidate image cache entries for file URIs. r=tn (31d73cb508) - Bug 1183563 - Fix incorrect mixed content warning after internal redirects. r=tanvi, r=seth (12a6c8a15f) - Bug 1150127 - Stop leaking windows via imgCacheValidator. r=baku (a7809c5fa7) - bits of Bug 1102048 (Part 20, imgLoader) (b2098c8a5a) - (No bug) - Correct blatantly lying comment in imgLoader.cpp. r=me DONTBUILD (64c42a5b09) - Bug 1160592 - Report image source size again in about:memory. r=dholbert (4e04cf3c3e) - Add an assertion for the first argument of NewImageChannel, no bug (4c8f087a8f) - Bug 1127534 - Remove assertion before creating a channel (r=sicking) (988692dc91) - Bug 1175371 - Make VectorImage wait to deliver LOAD_COMPLETE until its size is available. r=dholbert (3c81e0daff) - Bug 1181323 - Move nsSVGRenderingObserver's isupports/refcounting decl to subclasses, since one subclass (nsSVGFilterReference) already has its own redundant copy of the decl. r=dholbert (6171171c2c) - Bug 1161722 - If we're shutting down, don't warn about untracked unlocked surfaces. r=dholbert (f7e18ce481) - Bug 1170877 - Track how many times the SurfaceCache has overflowed and report it in about:memory. r=dholbert (884176cb1d) - Bug 1161743 - Upgrade 'WARNING: Not expiration-tracking an unlocked surface' to an assertion. r=dholbert (9900169e7b) - Bug 1167557 - Crash when a null surface is passed to SurfaceCache::Insert. r=dholbert (b3c4cf60aa) - remove bypass cache not fonud either in FF nor TF (3ed4056a27) - Missing bit Bug 1102048 (Part 25, header guards) - Make image/src files comply (cb8ed2428f) - No bug - Remove obsolete comment in SourceBuffer.h. r=me (6e9c233448) - coding style (94b7269690) - Bug 1157065 - GFX: 2D: Add Loongson3 MMI helpers. r=jrmuizel (ebce946c91) - reverto to FF52 and TFF settings (e147a8c7b5) - Bug 1134599 - Fix rpi build target. r=jrmuizel, r=shuang (b9722f860c) - Bug 1129147 - Part 1. Take CanvasPath into a separate file, to avoid circular dependency. r=roc (859bcad807) - Bug 1129147 - Part 2. Path option to addHitRegion. r=ehsan r=gw280 (b2ab08a8a1) - Bug 1206076: Use a specialized PersistentBufferProvider for Canvas2D when using a SkiaGL DrawTarget. r=jrmuizel (859589caf8) - Bug 1188752 - Addendum: Make PersistentBufferProviderBasic constructor explicit. r=bustage on a CLOSED TREE (a27a4dc974) - style (72a65dcb26) - Bug 1198574 - Remove unnecessary argument for PersistentBufferProvider. r=bas (dca718bba8) - Bug 1163124 - The initial value of the canvas filter property should be "none". r=roc (59df6a01d8) (e7535dc94b)

Yes and thanks...just got it.

Somebody from another forum has release an unofficial build of OpenVPN 2.5.4 for XP on Sourceforge: https://sourceforge.net/projects/openvpn-for-windows-xp/files/ Regards