Leaderboard

New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20250726-3219d2d-uxp-7e0ddfef73-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20250726-3219d2d-uxp-7e0ddfef73-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-20250726-3219d2d-uxp-7e0ddfef73-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-20250726-d849524bd-uxp-7e0ddfef73-xpmod.7z Win32 IA32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20250726-d849524bd-uxp-7e0ddfef73-xpmod-ia32.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20250726-d849524bd-uxp-7e0ddfef73-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20250726-d849524bd-uxp-7e0ddfef73-xpmod.7z Win7+ x64 AVX2 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20250726-d849524bd-uxp-7e0ddfef73-w7plus-avx2.7z Official UXP changes picked since my last build: - Issue #2691 - Follow-up: Add manual tests (5869164e31) - Issue #2826 - Set up fontconfig on XRE startup (GTK only) (b992da05eb) - No issue - Adjust CSP URI reporting more closely with the spec. (0a8f50fafa) - [gfx] Hold a stronger reference to gfxFontEntry. (40f96b82ad) - [js] Align br_table limits with V8/Gecko. (3f297457d8) - [network] Stop accepting nameless cookies. (cdc8127cd1) - [DOM] Make object loading use an explicit scheme whitelist. (f902d2ad0e) - Issue #2828 - Part 1: Build and store the array of layer name tokens in the layer at-rules (53e6317bfb) - Issue #2828 - Part 2: Collect and hold rule cascade data per cascade layer (7b51aa7ca1) - Issue #2828 - Part 3: Refactor selector matching and rule cascade data into separate files/classes (03f29e75ec) - Issue #2828 - Part 4: Move cache key storage to ResolvedRuleCascades (aab9cdf9bd) - Issue #2828 - Part 5: Implement child rule processors (e05d4d8306) - Issue #2828 - Part 6: Assert rules were added to the rule cascade data (d3b305b819) - Issue #2828 - Part 7: Walk child rule processors separately and ensure inverted cascade layer order is enforced for important rules (ec57be653a) - Issue #2828 - Part 8: Extract RuleProcessorGroup (formerly ResolvedRuleCascades) into separate file (3605630f4e) - Issue #2828 - Part 9: Rename CascadeEnumData to CascadeLayer (1655a5a4a9) - Issue #2828 – Part 10: Clean up cascade layer constructor (a922ec0823) - Issue #2828 - Part 11: Detach weighted rule data from cascade layer (cbb019a862) - Issue #2828 - Part 12: Implement cascade layer SizeOfIncludingThis (e3476d33ed) - Issue #2828 - Part 13: Exclude strong layers-related code from release builds (318031c74b) 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: partly revert rev 5e8e120 for keeping -moz-hidden-unscrollable to work (44e5bd7711) - [Basilisk] css: replace -moz-hidden-unscrollable with clip (c466a1385f) - [Pale-Moon] css: replace -moz-hidden-unscrollable with clip (9e02324b0b) 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.

r3dfox 140 will come soon

New security patched release. Supermium 132.0.6834.226 R5.01 "The only substantial change in this release is a patch for vulnerabilities CVE-2025-8010 (CVE-2025-8011 is not applicable to M132) and CVE-2025-6558."

Done. 😽

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-20250726-c41cda278-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk55-win64-git-20250726-c41cda278-xpmod.7z repo: https://github.com/roytam1/basilisk55 Repo changes: - import from `custom` branch of UXP: [Basilisk] css: replace -moz-hidden-unscrollable with clip (a0f628da2) - import from `custom` branch of UXP: layout: partly revert rev 5e8e120 for keeping -moz-hidden-unscrollable to work (9ce6fba76) - import from UXP: Issue #2691 - Follow-up: Add manual tests (5869164e) (d4cd9cc60) - import from UXP: Issue #2826 - Set up fontconfig on XRE startup (GTK only) (b992da05) (207207241) - import from UXP: No issue - Adjust CSP URI reporting more closely with the spec. (0a8f50fa) (96316a8d7) - import from UXP: [gfx] Hold a stronger reference to gfxFontEntry. (40f96b82) (b6584ea5e) - ported from UXP: [js] Align br_table limits with V8/Gecko. (3f297457) (720ef774c) - import from UXP: [network] Stop accepting nameless cookies. (cdc8127c) (15f4be2b2) - import from UXP: [DOM] Make object loading use an explicit scheme whitelist. (f902d2ad) (da9ac104c) - import from UXP: Issue #2828 - Part 1: Build and store the array of layer name tokens in the layer at-rules (53e6317b) (21ff29fca) - ported from UXP: Issue #2828 - Part 2: Collect and hold rule cascade data per cascade layer (7b51aa7c) (37d446504) - ported from UXP: Issue #2828 - Part 3: Refactor selector matching and rule cascade data into separate files/classes (03f29e75) (eb52d9cd2) - import from UXP: Issue #2828 - Part 4: Move cache key storage to ResolvedRuleCascades (aab9cdf9) (f6755c8be) - ported from UXP: Issue #2828 - Part 5: Implement child rule processors (e05d4d83) (595bed820) - import from UXP: Issue #2828 - Part 6: Assert rules were added to the rule cascade data (d3b305b8) (5cde0ecef) - import from UXP: Issue #2828 - Part 7: Walk child rule processors separately and ensure inverted cascade layer order is enforced for important rules (ec57be65) (7c4314376) - ported from UXP: Issue #2828 - Part 8: Extract RuleProcessorGroup (formerly ResolvedRuleCascades) into separate file (3605630f) (0a5d13c19) - import from UXP: Issue #2828 - Part 9: Rename CascadeEnumData to CascadeLayer (1655a5a4) (cbf54b1f1) - import from UXP: Issue #2828 – Part 10: Clean up cascade layer constructor (a922ec08) (28c6452d7) - import from UXP: Issue #2828 - Part 11: Detach weighted rule data from cascade layer (cbb019a8) (4b366e590) - import from UXP: Issue #2828 - Part 12: Implement cascade layer SizeOfIncludingThis (e3476d33) (40fa28911) - import from UXP: Issue #2828 - Part 13: Exclude strong layers-related code from release builds (318031c7) (c41cda278)

New build of BOC/UXP for XP! Test binary: MailNews Win32 https://o.rthost.win/boc-uxp/mailnews.win32-20250726-7bda12e7-uxp-7e0ddfef73-xpmod.7z BNavigator Win32 https://o.rthost.win/boc-uxp/bnavigator.win32-20250726-7bda12e7-uxp-7e0ddfef73-xpmod.7z source repo (excluding UXP): https://github.com/roytam1/boc-uxp/tree/custom * Notice: the profile prefix (i.e. parent folder names) are also changed since 2020-08-15 build, you may rename their names before using new binaries when updating from builds before 2020-08-15. -- New build of HBL-UXP for XP! Test binary: IceDove-UXP(mail) https://o.rthost.win/hbl-uxp/icedove.win32-20250726-id-656ea98-uxp-7e0ddfef73-xpmod.7z IceApe-UXP(suite) https://o.rthost.win/hbl-uxp/iceape.win32-20250726-id-656ea98-ia-c642e3c-uxp-7e0ddfef73-xpmod.7z source repo (excluding UXP): https://github.com/roytam1/icedove-uxp/tree/winbuild https://github.com/roytam1/iceape-uxp/tree/winbuild

I don't see why it would be, they are different exploits. I'm sure win32ss knows what needs patching and what doesn't.

Now, after @feodor2's versions cleanup, I was able to install Scratchpad 0.7.1 in Mypal 74.1.0 as a replacement for the removed, internal Scratchpad but it doesn't work as intended. The Scratchpad window is partially broken. No possibility to run JavaScript code. From this I conclude that Mypal 74 is not yet at the level of FF 72 or even 74 when it comes to extensions. @feodor2, am I right?

Today, I have installed and tested Mypal 74.1.0. @feodor2, thanks for this new version and for cleaning up the version number chaos! Good job! My favoured Alice0775 script loading method still works with this browser version as I have predicted in previous posts. And I am very pleased that the Custom Buttons extensions and all my self-created scripts also work there. However, one issue is still present in Mypal 74.1.0. Clicking entries in the Add-ons Manager does not work yet and leads to the well-known message that the address is not understood.

No, I don't, but thanks. I use VLC.

I read @D.Draker killed his B85 mobo with those originals, not sure about the fixed. What exactly did you fix?

Hi nicolaasjan, I tried to download this video https://www.youtube.com/watch?v=NGPeK8gO0Eo on Win XP SP2 with your yt-dlp Python 3.11 build from https://dl.dropboxusercontent.com/scl/fi/2ajuzjxmmbwqk888dj9hg/test_Py3.11_XP_yt-dlp.zip?rlkey=ro7rfvve4ueuavheq30ddvhu6 And I can confirm the same error you got: "Unable to download API page, Failed to establish a new connection: [Errno 0] Error"

这个我以前也注意到了，不过这个问题实际是Windows简体中文翻译或本地化的问题，因为sab实际引用的是系统日期时间控制面板的资源（timedate.cpl），具体为'@timedate.cpl,-1189'。 这个Second（应为“秒”）翻译成 第二个 的翻译错误应该是自Windows7以来就有了，估计可能是界面上没有吧，所以一直没人发现 。 同时上面的Day好像也应该翻译成周（week）而不是日（天），这个也是引用的（'@shell32.dll,-34830'）——或引用的不对。 那么怎么解决呢，要么作者调整下引用或修改添加到字符翻译里，要么自己修改系统的资源文件或sab资源。 我不知道其他语言（翻译）情况，看起来没人反馈吧，估计可能就是简体中文有这个问题。

1) new patchk32.bat handles this. 2) added but untested

Given that the yt-dlp issue tracker will only accept issues accompanied with a verbose log from their own release builds, I'm afraid any issues experienced with my build cannot be discussed there. Also, any issues with my fork will mostly also happen with the upstream build. So, first search their issue tracker to see if someone has experienced something similar. If nothing is found, then perhaps ask right here? Or, like another forum member had done, PM me. He had an issue with some Polish site and I was able to reproduce it with the official yt-dlp build. Then I opened an issue there and it was solved.

... And now yt-dlp have updated their random UA string to report Chrome 132-138: https://github.com/yt-dlp/yt-dlp/commit/c59ad2b066bbccd3cc4eed580842f961bce7dd4a#diff-b6d91e8afdfc47fb2b59e37d22d3621b79d252bc5a8205edc2b80530dcceecc2

We may be talking about different things here regarding PAE. Standard PAE is enabled on my machine, and always has been. Until recently though, I had not experimented with 'PAE patches' to extend the amount of RAM available on my 32-bit system. I do now have one of those available as an option on boot, and it is possible that my access to the 3TB disk will fail if I use that, if I go over 2TB on the disk. I haven't actually tried that, but certainly with the standard PAE configuration, I went over 2TB on my tests without any apparent problems.

Definitely. ADB can be used to make or restore backups, and has access to some places that normal apps would not, so these backups can be more complete than those made by other methods. It can grant or deny various permissions. It can install or uninstall apps. It can explore directory stuctures, copy, move, rename or delete files within device storage or to/from your windows computer, change file permissions or ownership, or generally do most shell commands. It can remount read-only partitions as read/write (though you still need to consider how to deal with dm-verity on some of them like /system). It can reboot the phone into recovery mode or bootloader mode, useful if you want to change ROMs. It may be able to change itself to root mode, depending if Xiaomi left some config settings on that would allow this. It also has a "sync" function in addition to the backup/restore functions. I find myself using ADB at least once a month or so, mostly because that crippleware MTP mode does not always preserve file timestamps correctly, where using ADB does. I have not looked into Xiaomi in particular, but that arrangement sounds typical for getting a bootloader unlocked. While it is not directly rooting, it is typically the first step in either rooting or in installing a custom ROM. One thing to get clear about before you do that, is for some vendors, requesting that unlock will void whatever manufacturer warranty the phone had, so best to check the rules before asking. My guess here would be that their web site is coded to want Chrome, and the needed functions are missing in whatever older browser you tried. If you don't like Google Chrome (I sure don't!) there are other chromium based browsers that are more privacy friendly. Brave ver 1.47.186, while not the most current, does run on W7 and is my go-to browser when my preferred SeaMonkey can't cope. Others might include Chromite, Vivaldi, etc; you can probably find a thread elsewhere on these boards about browsers still running in W7 for more ideas. Yeah, that is part of what I meant about the hassle factor. The card itself is usually pretty sturdy and not likely to break. But if your phone has one of those little pop-out trays that hold both cards, that thing is a lot more delicate and definitely prone to breakage, assuming it doesn't just get lost. I usually just install the card once and leave it in permanently, since ADB is simple enough for me to copy stuff out with. Yep. Even worse, for my FroYo device, doing this has about 50% odds of causing the phone to reboot, so I have to unplug everything, wait for it to finish, and start over. There's no real reason why the procedure to unmount and remount a partition should fail so often for androids when it is very reliable on Linux desktops, but apparently Google preferred to do away with it entirely rather than debug the actual problem here. Somehow desktop Linux systems solved this problem a long time ago and routinely stop trying to access a partition that is unmounted. I don't know why the same solution never made it into the android kernels. Oh, I just thought of another... if you have a USB-OTG adapter that fits your phone's USB port, you can plug in a standard USB thumb drive and use something like Ghost Commander to copy files in and out.

Yes, also on 8.0

for Server 2003 x86/x64 (Extended Security Updates) - KB4516665 (5.2.3790.6946) When you need multiple languages Here are all my Update Packs Source Files with update ESU that My friend and I kept it too have uploaded to archive.org https://archive.org/details/windows-2000-post-sp4-update-pack https://archive.org/details/windows-xp-post-sp1-update-pack https://archive.org/details/windows-xp-post-sp2-update-pack https://archive.org/details/windows-xp-post-sp3-update-pack https://archive.org/details/windows-server-2003-x86-update-pack_202506 https://archive.org/details/windows-xp-and-server-2003-x64-update-pack Another thing is Windows 7/Windows Vista/Windows 8/8.1 can use Windows Server 2008(NT6.0)/Server 2008 R2 (NT6.1) /Server 2012(NT6.2)/Server 2012R2 (NT6.3)Update/Patch because it is Code Base.

Thank you very much, MilkChan. I already used the file from Windows Server 2008 to verify whether the bug still exists. It does exist even in Windows Server 2025. But knowing which update contains the latest version of win32k.sys allows me to write an update for XP and Windows Server 2003, too. Otherwise I would be patching an old version which has other bugs. So, yes, I am interested in knowing the latest update which contains the win32k.sys for Windows Server 2003, too. At the moment I will stick with the 32 bit versions. I do not have much experience in writing 64 bit code. So I don't want to risk messing up someone's system.

just found UXP rev 5e8e1208 may cause some UI glitches, and workaround in both ways are committed.

You are on the right track, keep going. Edit : I notice also OTSKB.chm in Windows 10 require right click >> properties >> unblock to show the contains

@tinman2000 If you had read the last posts of this thread carefully, you would have noticed that Cloudflare does not work in Mypal 68 at the moment. No solutions, no suggestions. Maybe, one of the next Mypal 68 releases will successfully run these superfluous challenges. But I wouldn't count on that. Thanks to Cloudflare!

Webp is known to contain viruses, Chrome was patched, and was this codec patched? Nevertheless, not sure if it's a good idea to include it as a system wide codec. In details, below: https://msfn.org/board/topic/185031-webp-virus-fears-nightmares-suggestions-or-exodus-from-the-internet/

It says, it allegedly works on Windows 7 using the Ex. Kernel "Vx". And Vista is an absolutely different OS. And we can safely assume it won't work on a bare Windows 8.

Had a good laugh at the guy that claims to be from Southern Europe, writes Yandex is known for aggressive fingerprinting, yet tries to access it with such a zeal. Here's that extensive guinea pig test for either of you. Thank you and kudos to the tester. "Disabling all privacy-related extensions, including uBO" just to try to get to the favourite Yandex - wow! https://github.com/win32ss/supermium/issues/910#issuecomment-2362536383

Got the same.

Revo Uninstaller Pro is great. I have the last XP-compatible version 3.2.1 installed. If I had used this tool and not just that junk uninstaller which came with Wise Anti Malware, then I probably wouldn't have had that problem with the driver corpses and the resulting Thorium crashes. That's a very good idea. @Dixel Thanks again for pointing to the more recent NVIDIA graphics card driver in the version 370.90! This one is now installed in my native Windows XP Pro 32-bit system. Runs great.

I already modified the 360Loader.ini. I changed the name of both files to ThoriumLoader.exe and ThoriumLoader.ini and replaced the Icon Group of the ThoriumLoader.exe by the one from the thorium.exe. This modified loader works well to call up Thorium. And due to its separate ini file, the command line flags set by the user can be easily changed at any time. Here is a screenshot taken from my current "Open with" context menu in Windows XP: Everything I have written so far proves that Thorium can successfully take up and load a link via a command line call, initiated from the registry, when using a loader such as ThoriumLoader. Cheers, AstroSkipper

[Setup] AppName=Thorium [FileToRun] PathToExe=.\Chrome\Application\$AppName$.exe Parameters=--user-data-dir="$Profile$" --no-proxy-server /high-dpi-support=1 /force-device-scale-factor=1 --disable-component-update --use-angle=d3d9 --ignore-gpu-blocklist --single-argument [Environment] Profile=".\Chrome\User Data" This is what I actually meant. PS: Think of the Space Bug! Can you still remember?

That's very simple. Perform a right-click on a link to be opened, click on the item "Select a programme" from the "Open with" context menu and point to your 360Loader! When done, an entry is automatically created and stored in the registry under the HKEY_CURRENT_USER\Software\Classes\Applications\360Loader.exe key. From now on, you have an entry in the "Open with" context menu for taking up links by the 360Loader.

Ok. No problem.

Is this browser equipped with its own set of certificates? Looks like so.

@UCyborg You are absolutely right. MSFN uses definitely responsive design. I have installed a nice extension called Window Resizer and set up the same resolution as on my tablet. In portrait format, I now get the same, mobile view as on my tablet. Here is a screenshot with mobile view in Thorium under Windows XP simulating my Android 9 tablet: In this mobile view, MSFN websites called up in Thorium are more responsive and faster in loading than in desktop view.

No. It is indeed Thorium trying to connect to this address. To be more precise it is the process thorium.exe.

Yep! It works as it was designed. But personally I don't like this design. TZ Clock only shows the time and date when clicking onto the toolbar item. In contrast, my Clock custom button shows the time in the toolbar itself and when hovering, the weekday and date. But in the end, it's always a question of taste.

Astral Plain. No, I'm faster!

Coefficient Analysis

Actually, the About: Menu custom button should work in Serpent 52, too. When I am back at my desktop computer, I will have a look at this. In New Moon 28, however, I can definitely confirm that this custom button works as it should. Generally, most custom buttons tested by me should work in both browsers. But, of course, we all know the proof is in the pudding. I have just tested it and have unfortunately to confirm your observation. Although there is the compatibility string of Firefox {ec8030f7-c20a-464f-9b0e-13a3a9e97384} inside the code of the About: Menu custom button, the menu of all about: pages does not open in Serpent 52. The button was created many years ago, though. Therefore, it definitely was developed for much older Firefox versions (< 52). That's presumably why it is still working in New Moon 28. But there are still alternatives in the form of extensions for Serpent 52 which can be tested. I will have a look at that later.

As I already explained, I try to use as few extensions as possible, as they have a greater impact on memory and tend to cause compatibility problems. If features can be retrofitted by custom buttons, I use them instead of extensions. But I have to admit that I still install far too many extensions, but of course I also use them. As everyone knows by now, I'm also a big fan of extensions. The About: Menu custom button was created by the well-known developer LouCypher long ago. I just modded it a bit as described in my article.

Thank you very much! Confirmed fully working with Windows 8 ! You're a saviour! Topic can now be closed, thanks to @D.Draker.

Well done! Your Dutch colleagues will be delighted with such a service. BTW, I already mentioned this issue here two weeks ago: https://msfn.org/board/topic/185966-my-browser-builds-part-5/?do=findComment&comment=1258614 I have fixed my German language pack the day I posted this issue. This corrected language pack has been in testing mode since then. I will post it when testing mode is over, and all is definitely working. So far, no further issues.

But the biggest joke I've read there is that you can't download my release because there are adverts on Mediafire preventing the download. Where does he live?

Thanks for your report! But you are using the SSE version, and it can't be directly compared to mine. But maybe, I'm the only one who has the problem with the automatic start of the Browser Console. I'm curious what @roytam1 has to say about it. In any case, there is a problem with the Browser Console in the most recent version of New Moon 28 which wasn't there before.

Thanks, but I think it's not that script win32 wrote about, it looks like a custom made one.

"-In particular, the flags #disable-encryption and #disable-machine-id now make Supermium user data portable if enabled."

I'd like to read that article where "all security experts considered" this, thanks.

Either of you tried to mod the desktop drivers? There's a chance up to R430 Kepler might work. https://www.nvidia.com/download/driverResults.aspx/151873/en-us/

I'm now starting to suspect why we don't see win32 here often anymore.