Leaderboard

New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20251206-3219d2d-uxp-e4afc5655b-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20251206-3219d2d-uxp-e4afc5655b-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-20251206-3219d2d-uxp-e4afc5655b-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-20251206-d849524bd-uxp-e4afc5655b-xpmod.7z Win32 IA32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20251206-d849524bd-uxp-e4afc5655b-xpmod-ia32.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20251206-d849524bd-uxp-e4afc5655b-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20251206-d849524bd-uxp-e4afc5655b-xpmod.7z Win7+ x64 AVX2 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20251206-d849524bd-uxp-e4afc5655b-w7plus-avx2.7z Official UXP changes picked since my last build: - Issue #2489 - fix `color-mix` implementation. (103bd151c8) - Issue #2847 - Update tests and toolkit CSS to account for `windows-win11` (becf65d05a) - Issue #2869 - Make Purple Buffer additions check if cell is tenured. (34d1d5a650) 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: - layout: restore xp theme names in test (c80b9fc1ea) - ffvpx: update ffmpeg to 3.4.14 ( https://github.com/FFmpeg/FFmpeg/commit/9983d098ff0ee54bc3b77676dd885883bfbe4ffb ) (83cd56dfbb) - nss: update certdata and bump ckbi version to 2.82 (d8a03301d4) - dns: update effective_tld_names.dat from https://publicsuffix.org/list/public_suffix_list.dat (e27800bb2a) - icu: update timezone data (cb88bccc0a) - security: update EV root data (c417312184) - icu: import backport of ICU-20558 (Fix regression in DateTimePatternGenerator) (dc67193fba) - icu: if langTag is empty, point it to root.txt as well (9e8e5fbd83) - icu-data: regenerate with icu_sources_data.py (e4afc5655b) 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.

https://github.com/e3kskoy7wqk/Chromium-for-windows-7-REWORK He made few reworked versions. It also can support XP and Vista without extended kernels

I didn't realise there were so many still using SP2 (why?). I will consider building the next version with PyInstaller 5.13.2 again. But do note, that in the future that may not be possible any more, due to the changes they make. Consider upgrading to SP3. In the meantime, here is a build that should work on SP2. @davidz, Do you really mean you're on SP3? Because in my SP3 machine (VM), 2025.12.06.064237 works fine: [debug] Command-line config: ['-v'] [debug] User config "C:\Documents and Settings\Nico\Application Data\yt-dlp\config.txt": ['--rm-cache-dir', '--console-title', '--add-metadata', '--embed-thumbnail', '--convert-thumbnails', 'jpg', '--js-runtimes', 'quickjs', '--ppa', 'ffmpeg:-metadata synopsis=""', '-o', '~/Bureaublad/%(title)s.%(ext)s', '-S', 'res:1080,vcodec:avc,acodec:m4a', '--sponsorblock-remove', 'all', '-N', '6'] [debug] Encodings: locale cp1252, fs utf-8, pref cp1252, out cp1252 (No VT), error cp1252 (No VT), screen cp1252 (No VT) [debug] yt-dlp version nicolaasjan/yt-dlp@2025.12.06.064237 (winXP_x86_exe*) [debug] Python 3.11.4 (CPython x86 32bit) - Windows-XP-5.1.2600-SP3 (OpenSSL 3.5.2 5 Aug 2025) [debug] exe versions: ffmpeg N-121824-g775b102182-WINXP (fdk,setts), ffprobe N-121824-g775b102182-WINXP, phantomjs 1.9.8 [debug] Optional libraries: Cryptodome-3.23.0, brotli-1.2.0, certifi-2025.11.12, mutagen-1.47.0, requests-2.32.5, sqlite3-3.50.4, urllib3-2.6.0, websockets-15.0.1, yt_dlp_ejs-0.3.1 [debug] JS runtimes: quickjs-2025-09-13 [debug] Proxy map: {} [debug] Request Handlers: urllib, requests, websockets [debug] Plugin directories: none [debug] Loaded 1853 extractors Removing cache dir C:\Documents and Settings\Nico/.cache\yt-dlp ..

Those builds are originally for my private use and they are released because of my goodwill that I think they may help others. This bug doesn't affect me and so I don't have big intention on hunting and fixing it. If you want this to be fixed you at least need to find when the problem started to occur. and last but not least, stop gaslighting.

Maybe in Explorer options you have hidden system files and extensions.

Sorry, I can't see the mentioned files in your folders. I see only the file wsus3setup.cab in first screenshot but in second one none of them. And don't hide your extensions of files. They are important to have a look at.

Error code 0x8007000D means Error_invalid_data. Therefore I think using the XP version won't fix it probably. You have to find another version if it exists at all. Did you copy wsus3setup.cat too? Without it won't work. Both files wsus3setup.inf and wsus3setup.cat are included in cab file wsus3setup.cab.

@maile3241 I've checked the location of wsus3setup.inf once again. The file exists in two folders, one copy is in folder WINDOWS\SoftwareDistribution\SelfUpdate\Default\ and the other in WINDOWS\SoftwareDistribution\WebSetup\. Both files have been accessed by WU. Maybe you missed one location.

Maybe the file version of wsus3setup.inf is incorrect. Search for a version related to Windows 7. Maybe within a normal Window 7 system. wsus3setup.inf is located in folder SoftwareDistribution. But I've never used MU web site in Windows 7 so it is only a guess.

On XP only GDI rendering is working. I will try to use DirectWrite. 139 on vanilla Vista SP2 works, but if you run without existing profile, it has some function missing in shell32.dll. Edit: DirectWrite from Supermium working! I can't get working hardware acceleration, even with disabled blacklist and D3D9... All api-ms-win-crt, p_advp32.dll, pwp_shd.dll, p_ole.dll, pwrp_k32.dll, ucrtbase.dll - that's the files needed for working dwrite.dll from Supermium.

Release: https://github.com/LordOfMice/Tools/blob/master/nvme9x.zip https://github.com/LordOfMice/Tools/blob/master/nvment.zip https://github.com/LordOfMice/Tools/blob/master/nvme2k-src.zip readme: NVMe driver for Windows 9x. Base: nvme2k 1.0.0.2 https://github.com/techomancer/nvme2k Changes: Port for Windows 9x, 64-bit LBA & Fixes 1. Generic - added SCSI 64-bit LBA commands processing (large drives support). Tested on Windows 9x ONLY - fixed controler shutdown code - speed up of command queueing and it is less CPU consuming - fixed enormously low speed or false positive errors on executing of large size requests - various code rectifications 2. Windows 9x specific - port with 2 bug fixes - install script nvme9x.inf written Install and uninstall as usual hardware, nothing specific.

@modnar PowerStrip 3.7x runs on WinNT 4.0 but requires a supported graphics card - I test PCIe NVIDIA 6200 TurboCache with DVI-HDMI cable. For the program to work, you need to delete or rename the nvgpio.dll file to nvgpio.bak I use NVIDIA driver 71-84-winnt4-international.exe. PowerStrip detects and write EEPROM EDID without any problems - unless it is write-protected, of course. It can only program 128 bytes EDID: P.S. Only PowerStrip 3.85 can program EDID 256 bytes but under WinXP - tested same NVIDIA 6200 graphics card but only with vga.sys driver. The NVIDIA driver cannot be installed !!!

Sorry for the misunderstanding, but I am looking for updates that I haven't listed, yet. There should be something like 10 or 20 different versions of win32k.sys. But I know only the 3 I listed.

For XP and Vista, you will also very likely need "VMware Tools 10.0.12". They can be found at the Internet Archive - BUT since they seem to also contain items in violation of MSFN forum rules, I feel it best to not "link" them here. Google/Bing/DuckDuckGo for "archive.org vmware player 12.5.9" then do "parent directory" browsing until you stumble upon "VMware Tools 10.0.12". Some people label them as "legacy" or "preVista". edit: have not used, I tend to stick with VirtualBox versions 5.1.22 and 7.0.26 as I have my system configured to use BOTH OF THEM at the same time. I use 5.x for everything except Win11 testing, which needs 7.x (maybe 6.x, don't recall) or higher. and on second thought, the archive.org links "should" be safe to post here at MSFN. according to Wikipedia, VMware has been *FREE* since mid-2024 (news to me, free is why I've always used VirtualBox instead).

Navigating from the link in post #1, FileHippo has versions 15, 12, 6 and many others.

Version 19 no longer installs correctly on a 26100.1 system. The taskbar is hung and keeps looping on a msg to hold down ESC to keep it from loading, but that doesn't stop the hang. Killed explorer and ran the v16 install and taskbar is back to normal. It seems v19 is not downward compatible anymore.

I really think it is time for you to stop festering roytam over your YouTube playback issues. Roytam has replied and you keep festering, festering, festering. @Dave-H - can you please add this to your "watch list"?

Maybe the problem is in the splitter. You have updated ffmpeg 3.4.14, the new version of ffmpeg will not necessarily work better than the old one. You might want to try older versions of ffmpeg from 2009. They may work even faster. Because the new ffmpeg is very much oriented to 64 bit and windows 10. The ffmpeg developers are sponsored by donations from google and microsoft related entities. Also, if possible, I recommend you look up the answer in Centaury's code https://github.com/Feodor2/Centaury/releases

It is possible that the SP3 you used for comparison was somehow corrupted. On my hardware, SP3 was indeed "faster" than SP2. I am not a "gut feeling" type of guy. I took quantifiable and repeatable measurements that led me to using SP3 over SP2. Hard facts that were repeatable. No "gut feeling", no "placebo effect". That's not to say that you are "full of it", lol, it is POSSIBLE that SP2 is "faster" on your hardware. Just saying that was not the case back in the day when I compared the two. One POSSIBLE factor is that I *never* install a Service Pack 'on top of' an existing installation - I SLIPSTREAM IT INTO MY INSTALLATION DISK. Then FORMAT and INSTALL brand new.

Hi nicolaasjan, I still use XP SP2 for now, I tested the latest yt-dlp 2025.12.06 and when I try to download a video it says: "... api-ms-win-crt-convert-l1-1-0.dll was not found." hmm, it doesn't work with any command I try. (example: yt-dlp --version ) Can you please keep yt-dlp be able to work on XP SP2 if possible? I would also like to say a big THANKS to you nicolaasjan for all your hard work with yt-dlp and making video download possible on older OSs.

New build of post-deprecated Serpent/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 52.9, but some security patches can not be applied/ported due to source milestone differences between versions. Test binary: Win32 https://o.rthost.win/basilisk/basilisk55-win32-git-20251206-23f410065-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk55-win64-git-20251206-23f410065-xpmod.7z repo: https://github.com/roytam1/basilisk55 Repo changes: - import from UXP: Issue #2489 - fix `color-mix` implementation. (103bd151) (1791c6a30) - import from UXP: Issue #2847 - Update tests and toolkit CSS to account for `windows-win11` (becf65d0) (bb4464dfb) - import from UXP: Issue #2869 - Make Purple Buffer additions check if cell is tenured. (34d1d5a6) (be165adf3) - ffvpx: update ffmpeg to 3.4.14 ( https://github.com/FFmpeg/FFmpeg/commit/9983d098ff0ee54bc3b77676dd885883bfbe4ffb ) (c182ca3c9) - nss: update certdata and bump ckbi version to 2.82 (db7a51891) - update timezone tld and ev root data (58e09f3c7) - icu-data: regenerate with icu_sources_data.py (868997718) - icu: import backport of ICU-20558 (Fix regression in DateTimePatternGenerator) (72c196bdb) - icu: if langTag is empty, point it to root.txt as well (23f410065)

I appreciate your work, so no offense. I downloaded your first version basilisk52-g4.1.win32-git-20180224-dc7ceccf8-xpmod it already has this defect. And to check it is enough even just to throw mp4 file on browser tab and change playback speed in player. While checking noticed that for example in basilisk52-g4.1.win32-git-20180224-dc7cececcf8 CPU load during normal playback is lower than for example basilisk52-g4.6.win32-git-20200725-4d76d4e-uxp-1e0bb1d35-xpmod. The difference is very small, 2-5%. In general I realized that you do not want to fix or can not, too bad.

@Goodwin @roytam1 I'm not a professional, I just compile Mozilla based browsers for my personal use and hope for the best. For instance, I'm trying to compile Firefox 13+ with Windows 2000 compatibility. One thing that sometimes (but not always) causes runtime execution failure is clicking on tools -> options -> applications which can cause an error in MSVCR71.DLL. Also, after reverting the commit that broke Windows 2000 compatibility, I had a minor problem that is caused by opening a new tab, and the message displayed on the new tab page was relating to an invalid address. That annoyance was caused by the commit for Mozilla bug 728429, which mandated ASLR on Windows. In order to find that, I had to cherry-pick a range of commits between the one that broke Windows 2000 compatibility, and the one mandating ASLR, then compile the source code a few times to narrow down my search. The point is, if a problem does not exist in one of roytam1's browser builds, then exists on a later browser version, the right thing to do is find out what the last good version, and the first bad version is, then it would be easier to find when the problem started, and what commit caused it. Also, it is possible a bug might not affect everybody.

You're welcome, as for Vista, the guy in the topic you linked, wrote it worked with the early i3 series. Pentium you mentioned is a newer chip.

It "seems" wrong. Last year, I had found a browser for Vista users, shared it here, which you aren't, so why posting? You are on Win10 as far as I can recollect, so, obviously it wasn't intended for you! And it was good for its time, not anymore. But probably the developer will read and listen. Losing D.Draker, losing the reading of his private e-mail, all of that would be a huge loss for him. So I hope he will read the post and finally update, and other Vista users will benefit, too. See how kind and caring D.Draker is?

The browser is actively supported with security patches, it didn't "die", no matter how you wish it to. It's just the problem of such rare engine updates, which is probably fine for users in China with their restricted and censored internet, but not for MSFN website with its super frequent updates.

No. https://msfn.org/board/topic/182660-intel-hd-haswell-drivers-modded-for-vista/#findComment-1198948

@Dave-H If your plan doesn't work, consider my strategy then. The steps are necessary for further steps I wanted to perform. And I don't know whether they work with an old Chrome as default. The Internet Explorer is automatically set as default browser by Windows XP if no other has been set by the user.

The installer cispremium_only_installer.exe inside the zip archive is definitely the original one. I already extracted it and checked its content.

You're welcome. BTW, you can vote for a new Mypal 68 SSE release in this thread here like @exogenesis did or on GitHub.

Mypal 68.13.8b release for SSE 32-bit: https://codeberg.org/Theodor2/Mypal68/releases/download/68.13.8b/mypal-68.13.8.en-US.win32.zip

First, try to forcefully set the full RGB mode. Second, 10-bit is only supported for their pro-grade cards, no wonder you see the black screen, even if the monitor supports it. Third, newer cards are tend to be overbright, so again no wonder there's a drastic distortion, they simply distort the vibrancy. Try lowering the digital vibrancy setting.

@Cocodile, thank you for the information! I had a chance to read the topic soon enough after you posted! A very interesting fact they mentioned, even VPN services are infested with that malicious AppMetrica Yandex API, hundreds of them. We'd have to assume the popular Browsec also has it. No wonder it has so many negative reviews and articles questioning its security and reliability. Such VPN services are definitely not advisable to use on iOS.

From my research, WD is simply not capable to produce disks of any capacity above 6TB. I have WD60EFAX (6TB SMR), it's fine, despite the new not so popular SMR tech. It's still too new to tell about the longevity.

I'm afraid not, because it would require backporting the whole shared memory API from Windows 10 build 1809 to windows 8.1. Better/easier/legally safer to build chromium from scratch (source), which is open, so anyone can try.

Def. update is a blob in the format they previously used, just added new known vulnerabilities, collected during these years. .>>> Could you please not use CAPSLOCK, at least not that often, thanks.

@mina7601 If I understood correctly, you did all from scratch, just for testing purpose?

And that's exactly what I prefer to do in most cases: manually as much as possible to gain full control. Unfortunately it's not possible in all cases. One can say a bit "old school" but I take it rather than a compliment. AIO packages can be great but they are not my favourites.

No new Easter egg from me! Either I've been just a bit too careless while writing and correcting or it is due to a kind of senility! Thanks for reporting again! I'll correct that immediately, of course! No, the fixed version of cacert Updater can be downloaded from first post in my thread in section Downloads: The changes are minor indeed and can be done simply by applying cacert Updater Fixed. BTW the AIO package has been created by @Windows7fan from MDL. Therefore maintaining of it isn't my job. And I totally agree, if actions are performed manually, users get full control because one knows what has been done and what has to be performed next. Very important if verifying of each step is necessary.

As I wrote in my guide, everything can be done manually. That's how I did it. By doing things manually you have full control of what you do and verifying is much more easier. Here is the changelog to AIO Update version 1.2.1: "Version 1.2.1 3/6/22: Updated cacert.pem to the latest version. This improves compatibility with most websites including Windows Update." This is a quotation taken from MDL.

I am so sorry! Truth to be told I really thought you are female because of your name Mina (read in forum "Introduce Yourself!") meaning "heaven". Once again, I have to apologize to you! Do not be angry! Appearently I am not a Sherlock Holmes. Wow! You are the only one who has found my easter egg! That means you have really read my guide carefully! Thanks for reporting! Good luck! Hope it will work this time! Kind regards, AstroSkipper

But does anyone still use IE6 in Windows XP seriously? IE8 is crap but IE6...No way!

My version of 11/12/2019 is probably the same that you have used from November 2019. Then it works offline with this version but using WSUS server + WUMT doesn't work in your system. Is that correct?

Thanks for upload of your Windows Update log! As far as I can see in you have used WSUS proxy method to get notification from systray shield. Is that correct? As mentioned above it works in my system too but that's not the problem actually. Only using HTTPSProxy switched on, no WSUS server. Is it then working too? I don't think so.

I don't , I just ignore such websites. They can go and screw themselves . Less money for them , it's as simple as that. In the end , what's their goal ? To sell goods or to track us ? And again , not a problem for the Dixel's starter (can change your fake ID with a given interval or not change at all , just allow to generate it once).

Not the case with Dixel's starter . It generates a new hardware ID and fake fingerprint each time your start your browser . I don't know about the one he gave you (which version) , but should also be doing the same thing . If not , there's a newer one. Try visit such websites and see what happens . I always get a captcha if I turn the advanced privacy setting on in the starter , and even if I solve that captcha , it asks me again because it can't make a proper fingerprint . As far as I know , from what he told me , they fingerprint you by your videocard first . And then by your driver , CPU , RAM size , system and vendor . All that should match with what the browser reports . No matches : captcha again .

http://geocities.ws/jumper/

There you have it ! You need ONLY 64 bit . I will share this direct link to ALL official Opera versions from their archive . https://get.geo.opera.com/pub/opera/desktop/ You just need to extract it with 7zip , replace the files in your portable version, no installation needed . And use that portable launcher. Everyone , notice he has Aero working with Opera...

Well , it is already enough to run Opera . Maybe if you have some spare time and don't mind adding just shell32 sooner , with this function added, I'll start testing Opera with my starter. It's no hurry , of course , I have plenty of others to test . Today I've tested latest Slimjet - working , of course with this flag.

You also forgot Toshiba ! This is my list of quality/compatibilty/preference PCs and laptops with Vista (I've had experince with each brand) 1. Siemens PC and laptops (later became just "Fujitsu") 2. Philips and Siemens monitors (I mean in the old days , good quality and brilliant picture) 3. Toshiba notebooks 4. Sony notebooks 5. Dell notebooks (Win2000 - Vista era only) and some of their expensive monitors with Samsung screen (from 2006 - 2007) Everything else ..... hmmmm .... nah ....