Leaderboard

Not true, no additional items. The GUI is clean. Screenshot below.

Download p4. jisu9.zip.004

part 3. jisu9.zip.003

No problem, part 2. jisu9.zip.002

as promised .heic is brought to windows xp closing the gap for xp having a very new image encoder and decoder (and nothing using any other modules, engines or weird operating system dependencys) the resulting image is even better then the one on wikipedia https://en.wikipedia.org/wiki/High_Efficiency_Image_File_Format reason behind this i only choosed the best options, better internal code, better decisions, and disregarded code that decreases image quality ----- APP "WinXP HEVC/HEIF/H265 Image En/Decoder explained" The Encoder: Encode By Filename: allow you to select a file this heic encoder supports : .jpg, .png, .tif and .y4m (raw YUV format) the encoder make a .heic file from the choosen file Encoder By Folder: this read "Encode By Folder Searched Format Ending" if you have choosen jpg then it will search all .jpg files in the choosen folder actual chooses: png, jpg, tif or y4m if the entry was jpg then the encoder will encode all .jpg files in that folder to .heic this by folder was made so you can encode many files instead of always 1 file, while going to drink some coffee Encoder Complexity: "placebo" is the best setting here i actually dont see any reason to choose a different setting that actually only results in less image quality (best: 1: placebo, 2: veryslow, 3: slower, 4: slow, 5: medium, 6: fast , 7: faster, 8: veryfast, 9: superfast, 10 ultrafast) why we should we set a setting that decreases image quality? (basicly i not even wanted to have this box) (1 word about this, placebo use the best compressions tricks therefore the time is longer(more code = more time), the others try to speed up this (and leave out some functions, tricks, try to end the encoder before it really was done), it might not always increase the image - but you can be certain you got the best option (and the highest amount of pixels possible) also it can happen that that your image might not challenged the encoder for its maximum, then a "faster" setting dont have that much difference, still it can result in less amount of pixels, with placebo you are certain to get the maximum it really raise questions to use the others, you can make a big jpg file and you may dont see the image difference that much - but why ? are we making a jpg or are we making a high efficienty image encoder ?) Hardware Acceleration: makes use of hardware registers such as MMX, SSE and AVX this speed up encoding time a lot since the encoder is very complex image encoding can take time hardware acceleration makes encoding a lot faster notice: depending on your cpu power since the encoder is complex can take some time (if so keep a look "Encoded Image Files") (MMX, SSE and AVX are speed hardware registers they are between 64 and 512 bits wide, depending what one is available (yes in 32 bit)) Quality: controls the filesize of your .heic file, the lower this number the smaller your .heic file lossless: that option is not very useful as the real question is how well the pixels was preserved making a compression (we actually dont make a raw format - we make a compression) so better set this option to 0 (it dont make a real compression) Tuner: this increases the image quality even further good settings are psnr and ssim - the other settings only decrease image quality the tuner increased the amount of pixels, as said before its a extra function to improve more pixels more code = more time - this makes a good example - if you leave out many of good possible tricks you might end up in a less fancy picture ----- The Decoder: you have to choose a output image format for your .heic file (we have png, tif, jpg, and y4m) png compression level (0-9): png compression level -1 actually represents png compression 6 i actually dont see a well reason to have -1 as option, since -1 just represent compression 6 (you can try this out by looking at the filesize of the resulting .png file (try -1 and 6 they are the same) -1 actually is called png_default_compression what then is defined as 6 0 means no compression (this is good to make a compare how well your .heic file was preserved) increasing values make higher compressions losing more pixels (again 6 is equal to -1) 0 is the best png compression regarding pixels going from 0 to higher numbers decreasing image quality (higher numbers create smaller file sizes) (and make a compromise about pixels and compression) png is said to be lossless, but i only know for certain if option 0 is selected that it is a lossless copy. (what makes a 1:1 copy of the .heic file as it exits) jpg compression (1-100): nothing much to say here the higher this value the better the resulting jpg image notice higher values also cause bigger filze sizes 90 seems to be a good choice Decode By Filename: this button actually reads out "Decode In Format" why ? because if you select a .heic file the decoder has to know the decompression format valid formats are: png , jpg , y4m, tif Decode By Folder: reads out "Decode In Format" then the selected folder is searched for .heic files and then the decoder decodes all .heic files into the image format you set in "Decode In Format" Multi-pass: this makes a second image and compare the result with the first image - according to information this also improves image quality a bit (the h.266 says for example 1-3 % in average, then something about maybe sometimes more) Create A Subfolder: this allow you to put a folder where the WinXP HEIC en/decoder put its files it trys to create that folder, but you also can create that folder yourself this also avoid the name problem when controlling with "By Folder" ------ rumors say .heic is the best image encoder at the moment as we know .heic passes jpg, jpg2000 and jxr (jpeg xr) what we can see in the wikipedia site for heic maybe .heic also other jpg formats like the jxs format (what is rather speed orientated then quality orientated) there are some (jpeg xt, jpeg xs, jpeg ls, jpeg xe, jpeg xl) (https://jpeg.org) if someone wants to makes the compares the h.266 by frauenhofer or jxl would be candidates to try or even the others ... i actually never seen a h.266 frauenhofer image yet (updated the links): https://www.file-upload.net/download-15405155/WinXP_HEIC.zip.html https://www.mediafire.com/file/g9t94vi3dr4gycl/WinXP_HEIC.zip/file

Simple to explain, I don't know about @Saxon, but I run the original non-modified version of 2044, because I can't run yours due to the Trojan Warning, after which the browser simply can't start. Saxon also seems to use the non-modified version of 2044, hence the screenshot.

Don't know anything about telemetry. I didn't notice anything bad. My limit is up, I'll ask @Saxon, if he could add the other parts.

This one needs to be extracted from all 4 parts. jisu9.zip.001

Sorry for multiple posts - size limits. skin.zip

Replace all files with these. iframe.zip

English Dark window 10 skin for this browser. en_skin.zip

I can explain it briefly: - Download a portable version of Opera/Opera GX (or an offline installer) - Download CFF Explorer, DWrite.dll (from Windows 10+) and 7zip - Extract Opera and put that DWrite.dll in the folder where the browser is located - Right click opera.dll (it‘s around 200 MB) and open it in CFF Explorer - Go to Import Directory -> Kernel32 -> Scroll down to DiscardVirtualMemory - Change DiscardVirtualMemory to VirtualAlloc and save it - Go back to the folder where Opera is located -> Right click opera.exe -> Create a shortcut on the Desktop - Go to the Desktop - Right click the newly created shortcut -> Properties - On the General Tab write —no-sandbox next to the location and save it Opera should now work and open

Hello, Dietmar, I don't think modern websites will allow outdated versions of WebGL anyways.

You should be able to manually replace this file with the widevine folder. I have not tried.

Canary 4767 https://startisback.com/StartAllBack_setup.exe

@NotHereToPlayGames Yessa, now all works Dietmar

Warning: this version predates the WebP buffer overflow bugfix by nearly two months. I tried running later versions (known to run on Win 8.1) with the help of VxKex, but had no luck. As a result, I'm sticking with 115 ESR. Of course, if someone's willing to download the source code, apply the WebP fix, and rebuild this nightly release, I'd have no objection to that.

That's helpful. I had forgotten that the WebP bug wasn't even publicly acknowledged until 9/11. That makes 117.0a1 even less desirable, even if it (technically) runs. Sorry for the OT; we now return you to your regularly scheduled @roytam1 browser coverage....

OT: I seriously doubt that Firefox Nightly 117.0a1 build you reference has the "WebP patch"; https://thehackernews.com/2023/09/mozilla-rushes-to-patch-webp-critical.html CVE-2023-4863 landed on Mozilla Firefox 117.0.1, this is the stable/release channel of the browser, coming after initial stable release 117.0; 117.0.1 was what's "affectionately" called by Mozilla a "chemspill" release, https://wiki.mozilla.org/Release_Management/Chemspill i.e. the patch wasn't extant in the "upper" unstable channels (beta/dev, nightly) when it was first applied in their trees; at the time Fx stable was on v117.0, beta/dev was on v118.0b and nightly was on v119.0a1 - too late for deprecated 117.0a1 builds, which had served their purpose when the WebP vulnerability was made public... Last moment (before submission) addition: I located the post you mentioned: https://msfn.org/board/topic/185870-i-found-the-latest-build-of-firefox-nightly-that-works-on-windows-7/?do=findComment&comment=1253973 If you notice the Mozilla FTP link contained there, the Nightly 117.0a1 build you talked about has a build date of 2023-07-14, thus it doesn't contain the WebP fix, which was only applied on 2023-09-12 (two months after that build was compiled+uploaded) ...

later 2 are applied to my custom branch, structuredclone() of circular referencing object seems working (object taken from https://stackoverflow.com/a/1493501 for testing) I need someone to test chase with http://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20231021-3219d2d-uxp-dd9eca1b39-xpmod.7z and replace mozjs.dll with http://o.rthost.win/basilisk/mozjs52-custom-20231022-rev-88e66e9b.7z to see if it can solve the problem.

This is the main computer. which I use to write patches for xp/2k/2003/x64 Thank you for your help.

I just use the freebie v8.3.230608 and also use Ghidra v10.4 with Java v8 update 381. But I only use them to track down and isolate, I generally do all edits using HxD Hex Editor v2.4.0.0 x86.

Personally, I avoid Discord with a passion, because it really is unsuitable for old/under-resourced H/W like the one I'm currently on ... Additionally, all these Chrome-targeting chat services run sub-optimally on UXP, even on recent/supported H/W... Your main query has been answered here many a times, and (besides Discord) also encompasses similar services - Discord uses WebRTC for its audio and/or video features: https://sukhadanand.medium.com/how-does-discord-scale-to-5-million-concurrent-users-ed0874063fd but WebRTC, since the very beginning, was never supported by upstream in their PM browser (the same stands true for NM28); Serpent 52/55 do come with WebRTC enabled but, the last time I heard, their implementation is still eons behind the current iteration Discord etc. feel comfortable with... TL;DR: Discord are right from the start: Unsupported browser (you can still use text-based functionality, though) ... Since you appear to be on Win11, plenty of "workarounds" are available to you, just not for a UXP-based browser...

because it is hard to implement in current codebase: https://repo.palemoon.org/MoonchildProductions/UXP/issues/2229

StartAllBack 3.6.14

I turned off the warning, it still shown as insecure.

The padlock on your browser is not green, so it's not secure for you, too.

I get the same with ungoogled 110.

New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20231021-3219d2d-uxp-dd9eca1b39-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20231021-3219d2d-uxp-dd9eca1b39-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-20231021-3219d2d-uxp-dd9eca1b39-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-28.10.7a1.win32-git-20231021-d849524bd-uxp-dd9eca1b39-xpmod.7z Win32 IA32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20231021-d849524bd-uxp-dd9eca1b39-xpmod-ia32.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20231021-d849524bd-uxp-dd9eca1b39-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20231021-d849524bd-uxp-dd9eca1b39-xpmod.7z Official UXP changes picked since my last build: - Issue #2340 - Return an array (actually frozen sequence) for {border|content}BoxSize (9cd7fd3bf4) - Clean up dom\webidl (4ec7fa4f5f) - Issue #1442 Follow-up: Stop pretending proxies have a JSNative call/construct hook. https://bugzilla.mozilla.org/show_bug.cgi?id=1471924 Part 1 Also remove an erroneous debug assert and guard against future issues. This fixes crashes on vk.com but still does not behave correctly. (46024cd872) - PR #2347 - Backport of Mozilla tooltip bugs: (84ca5fdfc7) - Issue #1721 - Implement GlobalPrivacyControl (50ad087351) - Issue #1721 - Follow-up: Properly protect GPC header (0315aeee73) Official Pale-Moon changes picked since my last build: - [Pale-Moon] Issue MoonchildProductions/UXP#1721 - Implement GlobalPrivacyControl (591a887bf3) Official Basilisk changes picked since my last build: - [Basilisk] Issue #2 - Remove Safebrowsing and TrackingProtection from the FE (6cefc52d32) - [Basilisk] Issue #19 - Implement GlobalPrivacyControl (30e47eb5bc) - [Basilisk] Issue #19 followup - remove DNT from mozscreenshots We technically don't use mozscreenshots at all, but this should still be removed since it is no longer a valid setting (4edd2f91b4) My changes picked since my last build: - ported from mozilla: Bug 1446548 - Treat all GLSL versions >=450 as 450. r=kvark, a=RyanVM (43f0d1060e) (f508d00178) Update Notice: - You may delete file named icudt*.dat inside program folder when updating from old releases. * 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 don't remember when I said I don't use 360Chrome, you got me confused with someone else, nothing is hijacked since the Trojan message comes from the browser itself. Though, it's a bit different on XP and Vista. Did you have the chance to compare 1030 with 2044 with enough real life usage? Is it faster? Maybe not worth it? Or maybe worth to search for the cause?

Then maybe simply remove it? It wouldn't a problem, would it? I guess the browser will only be faster, no?

Yes, I have the same issue here. I'm on the Canary Channel though.

What I posted is to disable MS Updates not Defender. You have to use another antivirus anyway if you disable updates since Defender requires MS updates enabled to download definitions.

The following equation applies: ηMatrix = eMatrix (true) BTW, η is a Greek letter and not the Latin letter n. As @RamonUn already described above, uBlock Origin is a mighty tool when working in advanced mode. But it is more complicated in this mode than in standard mode. And misconfiguring is also much easier. uBlock Origin and ηMatrix combined gives the user full control over a website while surfing it. Simply great!

Such comments should actually be beneath you. Here, in this thread, it is about antivirus programs for Windows XP and not about personal sensitivities. Malwarebytes Premium 3.5 didn't have the top score results in the tests at the time, but it does work exactly as you configure it. No such problems with Malwarebytes as I have observed with Norton & Co. Here is a better overview and shows how Malwarebytes has developed up to the latest (not XP-compatible) version, which has very good scores, though: https://www.av-test.org/en/antivirus/home-windows/manufacturer/malwarebytes/

Even though I have a clear and incontrovertible opinion of Norton, there may well be others who prefer software that makes every decision for them. All I can say is: to each their own! And of course the population is larger in the USA (than in the UK), but the English language was only imported. It originated somewhere else. But honestly, I mix British and American English very often, rather unconsciously. Don't forget, I am not a native English speaker, either!

I actually thought that Norton had stopped any updates years ago. But no problem. I never liked Norton (Symantec or whatever) Antivirus & Co and felt they disenfranchised the user. So, rest in peace, Norton!

Yep, you can have a look in my new thread where I've started a current list of working antimalware programs with all necessary links and information which is maintained and extended continously by me. Kind regards, AstroSkipper

If the login is necessary for activation of KFA, then you will probably stuck without any further features like for example real-time protection. In terms of Bitdefender, the same problem occurs. No login possible, no activation. Fortunately, I am not particularly interested in these programs. And one thing is clear: Kaspersky has removed all resources relating to KFA 19. So I am not at all surprised that it is no longer possible to log into the My Kaspersky account.

I provided the download link of my cleaned and re-repacked version of KFA 19 in this post for all who are interested: Cheers, AstroSkipper

From what I know , there were two versions of GT1030 (at least two). The first (original issue) had GDDR5 memory, the second more advanced piece of crap - GDDR4 memory - cheap gimped VRAM. And it was released after the 398.11 came out , hence the distortions you see, but it's just a hunch. So you still can't find win32 ? Why we just don't tag him then ? @win32, be a dear, please consult the guy regarding your driver, I already wrote him it worked for me with my GT1030.

This is actually very, very strange, your questions. The setup.exe is inside of the package. The moderators can download and see for themselves. The download link leads to the legit MS website ! Maybe you have the Kaspersky AV that is well-known for blocking the legit files from Microsoft ? Or: I'm just guessing here, you have downloaded another (not mine) driver made by win32 ? In that case, this question is to him. Or: Just a hint - you can install any driver by using only its .inf file from the device manager !

Instagram. I think I wrote before : 2205 becomes sluggish overtime (with big IG accounts). Even though I have 16GB of RAM.

Zero crashes with 1106 on Vista (youtube or whatnot) , but you *forgot* to delete 3 junk files . Sesafe.dll is the chinese gov. spyware/redirector , at least what I've been able to read at some translated pages. I didn't dig deeper myself , and I shan't , since I'm somewhat *reqiured* to prove my every word to a *wannabe moderator* guy (not pointing fingers at anyone, lol). Don't know (and don't want to know) what setdef.7z and webmsg are, but definitely not needed and the browser works without them just fine ! Overall , 1106 seems faster than 2206. And it works better with IG accounts. EDIT : And NO BSOD , like the previous poster have mentioned.

And what "reputation" would that be ?

I supplied you with the offsets and locations of master keys for all of those extensions for your "build" , what other "proof" is needed that they are there ? I have posted this publicly. BTW , you couldn't find them yourself.

Why ? Because they just called themselves "ungoogled" ?

Problems do not appear when running it on Vista too. I was under the impression that video hardware acceleration is just not possible due to XP lacking DXVA 2.0, perhaps this is why there were reports about all those yotube crashes on XP . I'm on Vista and I have none , with or without acceleration. WEb-Gl is a security hole , we are kinda trying to make a privacy oriented browser here...

I think I already told you we are from the same generation . https://en.wikipedia.org/wiki/Generation_X . I still have plenty of things from the glorious 80's ! But hey , is there a place left where no russians mentioned ? I mean , something like a russian free zone ? I recently returned from my vacation and it was d*mn hard to find a place without "russo turisto" ! lol. But you've just reminded me about this again : I still don't have ANY problems with the language you had mentioned earlier , even with your 2206 ! I only see English everywhere , I deleted your user data and started from zero. Again , English everywhere , including websites , my system locale is en-UK , these srings are from my profile data. And even if I change my locale to US , no problems at all too. Of course the strings are en-US after reboot. "system_web_app_last_attempted_language": "en-UK", "system_web_app_last_attempted_update": "13.0.2206.0", "system_web_app_last_installed_language": "en-UK", "system_web_app_last_update": "13.0.2206.0" "system_web_app_last_attempted_language": "en-US", "system_web_app_last_attempted_update": "13.0.2206.0", "system_web_app_last_installed_language": "en-US", "system_web_app_last_update": "13.0.2206.0" I would like to ask @XPerceniol if he has any language-related problems .

@Jaguarek62 , so I'm assuming you did overclock your CPU . What's your voltages /frequencies ? I need to have 4.3 GHZ (at least). I was looking for a mobo here , but they don't have anything older than 2020. So it will be a newer AGESA. What was the lowest version you were able to downgrade to ? Also , perhaps you will enlighten me on what's the difference between B450 and B450M ? Thanks EDIT : found a board with the ordinary B450 (not M) , the one from the video . I see you have experience with AMD , what CPU could you advise ? I need something that will reach 4.3GHZ with sane voltages .