Leaderboard

New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20260801-3219d2d-uxp-bd1ccc490e-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20260801-3219d2d-uxp-bd1ccc490e-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-20260801-3219d2d-uxp-bd1ccc490e-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-20260801-d849524bd-uxp-bd1ccc490e-xpmod.7z Win32 IA32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260801-d849524bd-uxp-bd1ccc490e-xpmod-ia32.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260801-d849524bd-uxp-bd1ccc490e-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260801-d849524bd-uxp-bd1ccc490e-xpmod.7z Win7+ x64 AVX2 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260801-d849524bd-uxp-bd1ccc490e-w7plus-avx2.7z Official UXP changes picked since my last build: - No Issue - Remove the half-life experiment in HTTP cache (da157b41a3) - [network] Add usrsctp TSN gap check. (105f231910) - [layout] Add corner-case correctness check in DottedCornerFinder::FindBestOverlap. (49c7e3445f) - [DOM] Stop accepting audio samples with too low sample rates. (5faa70e3aa) - No issue - CallUpdateWindow is [[nodiscard]], but we don't need to use the result here. (dbfe4bd5a6) - [parser] Associate image elements with form only if the form and the intended parent are in the same tree. (5ac81bf6ff) - Issue #2404 - Improvements to CSS aspect-ratio spec compliance (3905cc5bd5) - Issue #2404 - Fix aspect ratio sizing for grid items (8cf6c58858) - [image] Make sure the image was fully and cleanly decoded before finalizing its animation state as complete. (e252b3856a) - No issue - Use `strnlen` instead of `strnlen_s`. (5a765da5ca) - Bug 2029734: Flush before calling sctp_process_init() (44365a9dbc) - Bug 2049523 - Make ToJSValue for float consistent with double. (db7d9c4c76) - Issue #2404 - Split aspect ratio enum types into a separate header to fix compilation issues (d6022d3847) - Bug 2045742 (34cb4816c2) - [DOM media] Perform more thorough checks on video regions. (06e746865c) - Issue #3061 Follow-up: Remove network tickler prefs (0f9c761769) - [layout] Fix nsFrameSetFrame primary frame handling of replicated children. (e5b3bc7cf9) - [network] Fix CacheEntry::mDoomCallback data race. (9111049721) - Bug 2027349 - reject empty nickname in PK11_TraverseCertsForNicknameInSlot. (3d7dbc8fb9) - Bug 2027362 - bound usageCount in PK11_UnwrapPrivKey to keyTemplate capacity. (ec58fff3ca) - Bug 2029901 - Guard against keySize overflow in IKE PRF/PRF+ output sizing. (1ed628fe3a) - Bug 2027373 - guard space subtraction in ssl_CallCustomExtensionSenders. (97201f8758) - Bug 2027371 - rewrite labelLen bound in tls13_HkdfExpandLabelGeneral to avoid unsigned overflow. (7af0472879) - Bug 2027346 - validate encoded EC params length and tag in SECKEY_ECParamsToKeySize/BasePointOrderLen. (0bcb0f57ff) - Bug 2029807 - Guard padding read against empty output in SEC_PKCS7DecryptContents. (4cf35f7b33) - Issue #3153 - Make sure buttons report a definite size for flex containers. Second try, this time ONLY do it for flex container or grids. This should prevent layout problems in non-flex scenarios. (cd90cc0fd7) - Bug 2030563 - Validate CKA_CERTIFICATE_TYPE ulValueLen in nss_cert_type_from_ck_attrib. (2be3800e92) - Bug 2027325 - Reject empty nickname in PK11_TraverseCertsForNicknameInSlot. (1a603bfe9e) - Bug 2030102 - avoid leaving a dangling ss->sec.ci.sid on allocation failure. (5dd26d41f6) - Bug 2032110 - fix two integer overflows on LLP64 systems. (79a227055a) - Bug 2027364 - Set tail pointer to null in static slot lists when deallocating. (f167cb61ee) - Bug 2028954 - CERT_DecodeAVAValue — Integer Overflow in Output Buffer Sizing. (c4cf8cf6e3) - Bug 2029433 - Bounds-check wrap index in PK11_GetWrapKey to match PK11_SetWrapKey. (23dc9da219) - Bug 2027353 - clamp input.len to testString size in pk11_mergeSecretKey. (4ce8d28c03) - Bug 2043887 - guard against integer overflow in CERT_Hexify. (73995df3a6) 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: - nss: update certdata and bump ckbi version to 2.90 (e5c222be77) Update Notice: - You may delete file named icudt*.dat and icu63.dll 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.

StartAllBack 3.9.24

Thanks @Tangy. I would be interested to know your results. Just an update, I spoke again today to Panda's technical support people at AnyTech365. Again, really friendly and helpful people, who knew exactly what I was talking about. They are thinking, as many here do, that Panda have actually withdrawn support for XP. They agree that doing this by just breaking the legacy version of Panda Dome, is not a good way of doing this, especially for paying customers, and they are trying to find out whether this really is the case or whether this has been done accidentally and will be fixed, if indeed it's possible to do that. There has certainly been no change made to the system requirements on Panda's website, and the article trumpeting its XP compatibility claims to have been updated on July 1st. They have given me a number to ring for Watchguard, Panda's parent company, so I will try giving them a ring tomorrow to see what they say about this.

@Dave-HYes , I do confirm your findings. Panda is fully functional ! PS : Let us monitor the situation for a few days to make sure that this is the case.

OK, me again! I hope I'm not jinxing it, but Panda Dome seems to be working properly again now for me on XP. I hadn't had a stop and restart of the main service for a few days, so yesterday I removed my shortcut files exclusion. All still seems to be working fine. Can you confirm this @Tangy? Whether they have fixed it as a result of our reports or whether they realised the problem themselves, I do not know. I just hope now that it keeps on working! It does look as if XP support is still a thing for Panda Dome, at least for now! Cheers, Dave.

New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20260725-3219d2d-uxp-30427c0213-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20260725-3219d2d-uxp-30427c0213-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-20260725-3219d2d-uxp-30427c0213-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-20260725-d849524bd-uxp-30427c0213-xpmod.7z Win32 IA32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260725-d849524bd-uxp-30427c0213-xpmod-ia32.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260725-d849524bd-uxp-30427c0213-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260725-d849524bd-uxp-30427c0213-xpmod.7z Win7+ x64 AVX2 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260725-d849524bd-uxp-30427c0213-w7plus-avx2.7z Official UXP changes picked since my last build: - Issue #1803 - Convert std::*<T>::value/::type to _v and _t suffixes. (1dd094e90b) - Issue #3061 - Remove network tickler. (3312ebb226) - No issue - Exclude some files from the preprocessor that don't have directives in them. (4b77043d11) - Issue #3144 - Implement CSSStyleSheet replaceSync (05467e9db6) - [js] Guard against OOB string sizes in internal Printf (90952505ec) - [gfx] Add bounds checks and limit reverse scanning of ligature data. (07271921c9) - [gfx] Hold a strong ref during GetFontAt lookup. (0b8fd38fb3) - [DOM] Make parser update form owner properly (0cd2c3acf7) - [js] Initialize BaselineFrame return value slot in OnLeaveBaselineFrame. (a809f575d1) - [js] Clamp length of regexp captures array for dollar substitution. (f8982ebd3f) 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: - devtools: restore preprocessor usage (30427c0213) Update Notice: - You may delete file named icudt*.dat and icu63.dll 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.

Sorry for the delay in responding again. Nothing has changed since my last report, the program is working, with my folder exclusions applied, with only occasional stops and restarts of the main Panda service. I can live with that in the short term. I'm still hoping, of course, that a fix will be applied and it will all come good again. Nobody has any definite evidence that Panda have deliberately stopped their XP version from working properly, only speculation. I will wait a while and ring AnyTech365 again if I've heard nothing from them about this. They will be as anxious as I am to get an answer from the Panda developers about this, I'm sure, as they will need to be able to give accurate support information to anyone else asking about this problem! Assuming that they do have a direct channel of communication to them, and it would be pretty stupid if they didn't, hopefully it won't be too long a wait for an answer, one way or the other.

New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20260718-3219d2d-uxp-edd86c9a9d-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20260718-3219d2d-uxp-edd86c9a9d-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-20260718-3219d2d-uxp-edd86c9a9d-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-20260718-d849524bd-uxp-edd86c9a9d-xpmod.7z Win32 IA32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260718-d849524bd-uxp-edd86c9a9d-xpmod-ia32.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260718-d849524bd-uxp-edd86c9a9d-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260718-d849524bd-uxp-edd86c9a9d-xpmod.7z Win7+ x64 AVX2 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260718-d849524bd-uxp-edd86c9a9d-w7plus-avx2.7z Official UXP changes picked since my last build: - Issue #3122 - Stop using <base> element in json viewer. (25440b453a) - Issue #3129 - Add palette_neon_intrinsics.c to moz.build (6ba6120115) - Issue #3129 - Follow-up: Add aarch64 to platforms enabling Neon in libpng. (97a7730717) - Issue #3134 - Import missing libpng files for loongarch and plug them into moz.build (f3ed8ce4a4) - Revert "Bug 2033848 - Use non-default response ID for GTK file picker accept button. r=stransky" (7e4c152210) - Issue #2829 - Support revert-layer in CSS-wide parsing (cacd5b2d61) - Issue #2829 - Implement CSS revert-layer cascade rollback (07a651cb70) - Issue #2829 - Handle revert-layer variable fallbacks (8a63f0d1e1) - Issue #2829 - Use layer index naming for revert-layer metadata (121fa183a0) - Issue #2829 - Regenerate devtools CSS database for revert-layer (3ad8472edc) - Issue #3140 - implement URL.parse() (a5610624d0) No official Pale-Moon changes picked since my last build. No official Basilisk changes picked since my last build. Update Notice: - You may delete file named icudt*.dat and icu63.dll 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 got a response from Watchguard technical support. "At this time, we have not received any official communication indicating that Windows XP support for Panda Dome has been discontinued. Based on the information currently available, Windows XP continues to be listed as a supported operating system for the legacy Panda Dome version referenced on the Panda website. Based on the information you have provided, this issue appears to be related to **Panda Dome**, which is a consumer (home) product. While we understand your concerns regarding the recent crashes of **PSANHost.exe** on Windows XP, we are unable to investigate or provide support for Panda Dome through WatchGuard Technical Support." So it appears that they only deal with professional business customers at Watchguard support, and they can't help private domestic customers. I would have to use the domestic support channels for Panda support, which I've already done, of course! The guys at AnyTech365 have promised to let me know if they hear anything about the status of XP support from Panda, so all I can do is wait.

Assuming you haven't already uninstalled Panda, could you try excluding from scanning the folders I detailed in my earlier post? I would just be interested to know if this helps the problem on your system. I'm still getting occasional stops and starts of the main Panda service, and I'm trying to find out exactly what is triggering them. Again, not a fix, just a better understanding of where the mechanism is now failing. Incidentally, Panda's online support forms are pretty useless, as once you've submitted one, you are supposed to be able to add new information to it by clicking on the link in the e-mail it sends you. Unfortunately, this doesn't work. When you submit the revised form, you just get an error message that the e-mail and customer number don't match any current user! I've had to submit two further forms from scratch to give them more information.

Back again. I'm cautiously confident that I've found another workaround for the problem with Panda Dome, and this one appears to have stuck. I was puzzled as to why the crashes, although sometimes apparently at random, could be triggered at will, at least on my machine. One thing which nearly always triggered an immediate crash, even if the system had been running for some time without any, was to click the Windows 'Start' button. This led me to think that there was some issue with Windows Explorer which was triggering the crash. One other clue was that one of the first things I noticed was that if I clicked Start, even if it didn't crash, the user-added shortcuts were a blank box, only populating after several seconds. I tried excluding Explorer from scanning, but that made no difference. I then tried excluding my whole system drive from scanning. Hey presto, no more crashes! So, by a process of elimination, I eventually narrowed it down to the folders which were triggering the crashes when they were scanned. I now have three folders excluded from scanning – Documents and Settings\[Username]\Start Menu Documents and Settings\[Username]\SendTo Documents and Settings\[Username]\Recent So far this has stopped the regular crashes, and Panda has updated without them coming back again. This is pure speculation, but what I suspect happened here is that Panda rolled out an update which broke compatibility with the Windows XP Explorer, which is quite different from that on later versions of Windows, of course. When Panda tried to scan those folders, like the 'Recent' folder being scanned when 'Documents' is clicked on the Start Menu, this triggered a crash. This is of course only a workaround, not a fix, and the problem could come back at any time, but at least on my system it seems to have fixed it for now. There is a security issue with excluding system folders from scanning, of course, but as the folders concerned only contain shortcuts, I would think it's pretty minimal. If anyone else still has a malfunctioning installation of Panda Dome installed, I would be interested to know if this works for them.

FFmpeg update. XP: static shared libfdk-aac VISTAx86: static shared libfdk-aac The following static links will always match the latest uploaded XP archives: https://files.videohelp.com/u/69760/ffmpeg-win32-XP-static-latest.7z https://files.videohelp.com/u/69760/ffmpeg-win32-XP-shared-latest.7z

Regarding panda forum reporting, to be able to uninstall panda you have to be able to stop the antivirus running. right clicking task bar icon does not always respond. after a few reboots my task bar icon informed me that my computer was not protected, a tiny red circle with the x. then it was easy to uninstall from add/remove. I also used their generic uninstaller which removed everything from my pc. I did a clean install but unfortunately the song remained the same. 5 minutes later it crashed. I have not got the foggiest how to go about it.

MSFN’s gone completely haywire and is now even posting multiple times. It’s been nothing but glitches for weeks.

It Seems a Lot Like Dead Internet Has Actually Come True "Agentic AI activity is driving a lot of the browser activity you're seeing." By Frank Landymore Published Jul 28, 2026 https://futurism.com/artificial-intelligence/dead-internet-theory-come-true The infamous dead internet theory was originally a conspiracy that posited that the vast majority of the internet was run by bots and thinking machines. Most accounts you saw online were really bots masquerading as people - all part of some nefarious scheme to control humankind by keeping us isolated and divided. That last part is far-fetched, but the core of the idea - that bots outnumber humans online - is inexorably starting to feel weirdly grounded. Fortune did a roundup of the latest data on the makeup of internet traffic, and most of it points to humans now being a quaint minority. "There’s a lot of missing pieces of information, but a lot of the observed data suggests the same thing, which is like there is more bot activity," Rudy Yang, a financial tech analyst at Pitchbook, told Fortune. "Agentic AI activity is driving a lot of the browser activity you’re seeing." This should come as no surprise. Bots have been everywhere on social media for a while now, and more and more people are wielding little posses of AI agents to do stuff on the internet on their behalf, ranging from conducting research to shopping. OpenAI CEO Sam Altman infamously said he was worried that the dead internet theory was coming true last year, seemingly without a shred of self-awareness. Further corroborating this new paradigm, Cloudflare announced in June that 57 percent of webpage requests were being made by bots, marking the first time that bots made up the majority by its own metrics. Bots have long been ubiquitous, but AI agents are doing the most to reshape the web. One report from the cybersecurity firm HUMAN found that traffic generated by AI agents surged by 7,851 percent year over year, Fortune noted. All those AI agents aren’t generating ad impressions, which is how a lot of websites keep the lights on, while putting a strain on those websites’ servers. "They consume the web completely differently than humans do," Yang told Fortune. "It’s almost like an entire new category, customer category, was created, and it means a lot for businesses because no one, as a business owner, is going to want to silo themselves from being able to serve a completely new customer segment." They’re also changing what we see on the web, too. A report published in October from the SEO firm Graphite found that over half of all new English-language articles on the internet were written with AI. It hasn’t helped that Google, the internet’s de facto steward, has made a hard pivot towards making the internet less navigable by humans. Now, the focus is on having a chatbot to dredge up answers, instead of providing a list of links to websites they’d find in its search engine.

@Dave-H With exclusions in place, Panda has been working flawlessly. Tonight I'll remove exclusions to see what happens. Keep you posted.

I would like for you to replace the links for RogueKiller last XP compatible versions from generic web-site links to archived links with v15 and portable v16 on 32bit and 64bit. It's a shame that Panda is placing XP on guillotine right now, but to be honest, XP had a good run, 25 years of anti-malware support from various programmes. I am still using Avast on XP VMs and i feel relatively secure with COMODO Firewall Pro from 2007 version, and i have no complaints. I do not recommend the last version of COMODO Internet Security however, because it choked my old Celeron to the core when i tested 4 years ago...

# [Windows 8.1 x64] Cyberpunk 2077 2.31 Steam compatibility patch I am publishing an experimental compatibility package for **Cyberpunk 2077 2.31 on Windows 8.1 x64**, tested against the Steam public build `20383525`. ** Download has been taken down **ZIP SHA-256:** `5D8543335269D86B7456B60B6E6A3908C6C296B186DB93F8EF25EA38D371745E` > This release was tested only on **Windows 8.1 x64**, so that is the only > operating system for which it is currently recommended. The installer rejects > other Windows versions and unsupported game files. ## What it fixes The package provides a D3D12On7 raster path for the tested Steam build and fixes the known presentation problem where Cyberpunk appears frozen and updates only when another window is dragged over it. The game can be launched normally with Steam's **Play** button after adding the launch option printed by the installer. ## Performance On my tested machine, performance is close to running the game on Windows 10. It is dramatically better than my older Cyberpunk 1.63 compatibility setup, which delivered roughly half performance on the same machine. This is an observed real-world comparison, **not a controlled or standardized benchmark**. Results will vary with GPU, driver, settings and system setup. ## Tested setup - Windows 8.1 x64 - Cyberpunk 2077 2.31, Steam build `20383525` - GeForce RTX 2060 - NVIDIA driver `474.11` - 1920 x 1080 - Raster rendering only Other GPUs and drivers, GOG/Epic releases, hybrid graphics and other operating systems have not been tested. ## Required: official D3D12On7 1.1.0 package Download Microsoft's official [`Microsoft.Direct3D.D3D12On7` 1.1.0 NuGet package](https://www.nuget.org/packages/Microsoft.Direct3D.D3D12On7/1.1.0) separately and save the unchanged file next to `Install.cmd` as: `Microsoft.Direct3D.D3D12On7.1.1.0.nupkg` Expected package SHA-256: `C0C4D597537ECC1C777FA3B5523E5CCDE0B9984B88FB9F37B2385E1D687BF536` The Microsoft package is not bundled with this patch. ## Required: SecondSystem SecondSystem must already be installed and active. **It is not bundled with the patch.** Obtain it separately here: [SecondSystem from win32's Patreon](https://patreon.com/win32?utm_medium=clipboard_copy&utm_source=copyLink&utm_campaign=creatorshare_fan&utm_content=join_link) Please support its developer and follow the terms provided with it. ## Installation 1. Extract the complete ZIP. 2. Download the official D3D12On7 package above and place it beside `Install.cmd`. 3. Install and activate SecondSystem. 4. Exit both Cyberpunk 2077 and Steam completely. 5. Run `Install.cmd`. 6. Copy the exact Steam Launch Options value printed at the end. 7. Paste it into **Steam > Cyberpunk 2077 > Properties > General > Launch Options**. 8. Start the game with Steam's **Play** button. Do not append custom arguments; the included launcher intentionally uses only the fixed, validated command line. The installer validates the exact game/runtime hashes and stops before changing anything if the files are unsupported. It keeps all compatibility files local to the game directory and creates a verified uninstall record. ## Important limitations The following features must remain disabled: - Ray tracing and path tracing - DLSS - NVIDIA Reflex - Frame Generation, including XeSS Frame Generation This release is for raster rendering. Cyberpunk 2.31's newer Agility SDK path is not a Windows 8.1 target, so modern features that depend on it are outside this patch's scope. ## Removal and game updates Close the game and Steam, run `Uninstall.cmd`, then clear Steam Launch Options. The uninstaller restores any files it backed up and reverses its SecondSystem mappings. The package is locked to Steam build `20383525`. The launcher compares the full contents of the current executable with its validated local copy and refuses a mismatch. After a Steam update, uninstall this release and wait for a version explicitly validated for the new build; do not remove the version/hash safety checks. ## What is and is not included The ZIP contains the original open-source proxy, launcher, installer scripts, source code, documentation and notices. It does **not** include: - Cyberpunk 2077 files - Microsoft D3D12On7 or D3D11On12 binaries - SecondSystem - Steam files - any DRM bypass A legal Steam copy of the game is required. The installer extracts the runtime inputs from the official NuGet package separately supplied by the user and creates a local working copy. Microsoft terms still apply and may restrict this use or modification; this is not legal advice. Full installation, troubleshooting, technical and checksum information is in the included `README.md`.

New build of BOC/UXP for XP! Test binary: MailNews Win32 https://o.rthost.win/boc-uxp/mailnews.win32-20260718-40a79c75-uxp-edd86c9a9d-xpmod.7z BNavigator Win32 https://o.rthost.win/boc-uxp/bnavigator.win32-20260718-40a79c75-uxp-edd86c9a9d-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.win32-20260718-id-656ea98-uxp-edd86c9a9d-xpmod.7z IceApe-UXP(suite) https://o.rthost.win/hbl-uxp/iceape.win32-20260718-id-656ea98-ia-c642e3c-uxp-edd86c9a9d-xpmod.7z source repo (excluding UXP): https://github.com/roytam1/icedove-uxp/tree/winbuild https://github.com/roytam1/iceape-uxp/tree/winbuild

Long story short, I have revived Comodo Antivirus 12.0.0.6818 (CAV) as part of the Comodo Internet Security suite (CIS) since it was no longer working out of the box. It's one of the last working antimalware programmes with real-time protection, HIPS, VirusScope, secure containmenr, online definition updates, even incremental definition updates, and the possibility of offline updating. This is now my new antimalware programme. And Vir.IT of course. All free of charge.

As I already said, it is no longer time to be picky when it comes to real-time protection under Windows XP. And of course, I read such reports about IOBit Malware Fighter. I personally never liked the IOBit company. Their products have never really been any good. And in the case of IOBit Malware Fighter, I can see why. The positives: Easy to install. A clearly structured, very tidy and modern interface. Good usability. Lots of features not usually found in freeware. Excellent uninstallation speed. The negatives: No daily updates as advertised. Today, it’s still installing the version from 9 July. Not really integrated efficiently into Windows XP. When the protection is activated, performance plummets. Version 13 is constantly being promoted, and you’re told to download it, even though it’s not compatible with Windows XP. No self-protection. The scan speed is so poor that I’m lost for words... The scanning speed is the absolute worst I have ever experienced. Smart Scan was cancelled after 123,226 files!!! and 120 minutes!!! with only 5% progress!!! This is no Smart Scan. I then uninstalled it completely. The uninstallation was done in seconds. That just goes to show how poorly the programme was integrated into Windows XP. It was practically just a superficial add-on. And it can even be shut down completely with just two clicks. That’s a goldmine for malware. My conclusion: Looks great on the outside, but a flop on the inside. P.S.: And that’s the good thing about my ancient P4 computer. It mercilessly exposes the weaknesses of software.

@j7n are adamant on NOT using an extension-based adblocker (e.g. uBO) in the browser, because they harbour the conviction that such an extension would considerably impede their browser's performance ; in the past, both myself and other members here have suggested to them the installation and use of uBO, but they still think that their old ways of dealing with ads (blocking "known" ad-serving domains in their HOSTS file) is still sufficient/effective in the web-ecosystem of 2025+ ... Is this on a UXP-based browser (the object of these megathreads) ? Because if you're talking about recent Chromium/Firefox, these now come with built-in ways of dealing with (some of the) unwanted content ...

For the time being I am trying to avoid going back to a full build which may contain many libraries that nobody uses. It will expand as fixes and requests are added to it. Here is a build for you to test with added support for https. The next release can include it in all builds. https://files.videohelp.com/u/69760/ffmpeg-8.2-2237-b4b30ff8fd-win32-shared-xpmod-P4.7z

Thank you once more for these new releases ; however, I have to report that the latest ffplay.exe compiles for Vista x86, BOTH the static+shared binaries, are completely broken under Vista SP2 x86 (with ALL MS updates until EoS, and then some ); whatever media file you throw at FFplay.exe (e.g., via drag-n-drop) results in the app instantly crashing; below, an instance of a plain MP4 file containing h264 video + aac audio: Your previous VISTA release works as expected with that same MP4 file: FWIW, your latest FFplay.exe static compile for WINXP doesn't suffer from this issue and plays the MP4 file without any problem: You'll probably need a Vista SP2 x86 VM to troubleshoot this, as I have no clue as to what might have gone wrong; perhaps an updated third-party lib has in the interim (n8.2-dev-1568...n8.2-dev-2232) removed support for NT 6.0? We know it's not caused by upstream (FFmpeg source code), because your n8.2-dev-2232 compile for XP works OK; in addition, when I probe the MP4 file with latest FFmpeg for VISTA, it doesn't crash: ffmpeg version N-125505-gc57660fb18-WINVISTA Copyright (c) 2000-2026 the FFmpeg developers built with gcc 16.1.0 (Rev5, Built by MSYS2 project) configuration: --disable-autodetect --enable-amf --enable-bzlib --enable-d3d11va --enable-dxva2 --enable-iconv --enable-lzma --enable-zlib --enable-sdl2 --enable-gmp --enable-libmp3lame --enable-libopus --enable-libvorbis --enable-libx264 --enable-libx265 --enable-libdav1d --disable-debug --enable-libfdk-aac --enable-fontconfig --enable-libass --enable-libbluray --enable-libfreetype --enable-libmysofa --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libwebp --enable-libxml2 --enable-libzimg --enable-libshine --enable-avisynth --enable-libxvid --enable-libopenmpt --enable-libsrt --enable-libzmq --enable-libgsm --enable-chromaprint --enable-frei0r --enable-libaribb24 --enable-libbs2b --enable-libcaca --enable-libcdio --enable-libflite --enable-libfribidi --enable-libgme --enable-libilbc --enable-libmodplug --enable-librubberband --enable-libzvbi --enable-openal --enable-libcodec2 --enable-ladspa --enable-libuavs3d --enable-libjxl --enable-libxeve --enable-libxevd --enable-libharfbuzz --enable-liblc3 --enable-libaribcaption --enable-lcms2 --enable-libdvdread --enable-libdvdnav --enable-liboapv --enable-libqrencode --enable-libquirc --enable-opengl --enable-gnutls --enable-gpl --enable-version3 libavutil 61. 2.100 / 61. 2.100 libavcodec 63. 5.100 / 63. 5.100 libavformat 63. 3.100 / 63. 3.100 libavdevice 63. 2.100 / 63. 2.100 libavfilter 12. 2.100 / 12. 2.100 libswscale 10. 2.100 / 10. 2.100 libswresample 7. 2.100 / 7. 2.100 Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'Mastermind_Se8Ep12[2494504515559].mp4': Metadata: major_brand : isom minor_version : 512 compatible_brands: isomiso2avc1mp41 encoder : Lavf62.13.101 Duration: 00:25:41.93, start: 0.000000, bitrate: 968 kb/s Stream #0:0[0x1](und): Video: h264 (Main) (avc1 / 0x31637661), yuv420p(progressive), 768x432 [SAR 1:1 DAR 16:9], 837 kb/s, 25 fps, 25 tbr, 90k tbn (default) Metadata: handler_name : VideoHandler Stream #0:1[0x2](eng): Audio: aac (LC) (mp4a / 0x6134706D), 48000 Hz, stereo, fltp, 125 kb/s (default) Metadata: handler_name : SoundHandler At least one output file must be specified Hopefully you'll be able to get to the bottom of this, as I do use FFplay.exe myself under Vista SP2 (32-bit) ; thanks for all your hard efforts already!

You’re welcome! It’s rare for me to be this blunt. But what I went through last night was hell. My entire partition was at risk. The BSOD had wiped out various files. Fortunately, most of them were related to Panda. It was a miracle that I was even able to boot the partition at all. And the fact that Panda could be uninstalled despite the damage was another one. Incidentally, the whole drama started straight after the nightly automatic update that Panda had carried out yet again. I’m a bit reluctant to make recommendations these days, but I can tell you what I did a few days ago, in a moment of foresight. It’s really just a coincidence, but take a look at number 7 on the list you posted. That’s what I recently installed: VirIT eXplorer Lite. Here’s my post: https://msfn.org/board/topic/184730-antimalware-firewall-and-other-security-programs-for-windows-xp-working-in-2023-and-hopefully-beyond/page/101/#findComment-1288687 This antimalware programme is very nice. It is fully compatible with Windows XP, even in the latest version!!! The Lite version is free of charge. If you want to have more protection, you have to buy a license for the Pro version. The programme interface isn’t modern; it’s got a bit of a retro feel to it. That’s what programmes looked like in the 1990s: extremely colourful, with lots of bright colours and plenty of colourful buttons. Personally, I think it’s brilliant. It reminds me of those days. VirIT eXplorer Lite requires very little RAM and CPU power. The virus definitions are downloaded locally to the computer and are therefore available offline. That’s the only reason why the programme folder takes up just over 500 MB. And support is available and active. The intrusion detection shield, called VirIT Lite Monitor, had flagged processhacker.exe as malicious and issued a warning every time the computer started up. I then uploaded the file to VirIT or, to be more precise, to C.R.A.M., along with a message. A day later, the warning had disappeared, as they’d complied with my request to whitelist it. I’m completely satisfied with VirIT eXplorer Lite at the moment. Another reassuring point is that this is a programme developed in Europe – specifically in Italy. I will soon be writing a detailed article about VirIT eXplorer Lite/Pro. Old screenshot of my VirIT Lite Monitor with false positive: New screenshot with Process Hacker 2 whitelisted by C.R.A.M.: Cheers, AstroSkipper

For experimenting with Avast, you should avoid the last XP-compatible version. As far as I can remember, is a previous version associated with fewer problems. I have to look in my archive for the exact version number. As far as I can remember, there were repeated reports on the Avast forums that avast! Free Antivirus 18.5.2342 runs much better than the latest XP-compatible avast! Free Antivirus 18.8.4084.0.

All Panda Dome installers come with offline installer version 15.14.5.0. Therefore, this information is unfortunately of no use. I asked you for the programme version. The most recent is 22.03.05 but I currently use 21.01.00. And it doesn't matter whether you actively click the update button or not, the update is automatically done every day and there is no option to disable it.

The file version of the installer file is meaningless, unfortunately. It isn't the version of Panda Dome that it will install. It's probably just the version of the packaging process used. As far as I can see, the only way to find out what Panda version it is, without actually installing it, is to extract it with 7-Zip and look in the extracted \Program Files\Panda Security\Panda Cloud Antivirus\ folder. There you should find a file called version.ini, which, if you open it, will list the Panda version and build number.

Well, I guess of those here still using XP, very few use Panda Dome, even though it's one of the vanishingly few security programs which is still supposed to work on XP. I found one instance on the virtually dead Panda forum reporting the same problem (with no response). https://support.pandasecurity.com/forum/viewtopic.php?t=12151&sid=a73549f228b82be2dde1657a770a28e4 My installation is still crashing all the time. It stops if I clear the recent files in the cache folder but starts again as soon as the folder repopulates.

Yes indeed, in fact, the first test I did was to replace that entire folder with one from an ISO image I made a few days before the fault appeared. It made absolutely no difference, it still carried on crashing, so I discounted the contents of that folder as having anything to do with the problem. I next looked for any other recently updated files in the Panda installation, and that led me to the cache folder in the Panda program files folder.

I implemented in my ProxHTTPSProxy's PopMenu TLS 1.3 3V3 package the tool Min2Tray which I therefore have been using for years. So I can minimise any other programme to the systray as well. That was the best I could find.

Check now.

It is a built-in safety feature derived from older Firefox architecture designed to protect your system from unverified files and drive-by downloads.

User-Agent Switcher and Manager 0.5.0. You can get it here: https://addons.mozilla.org/de/firefox/addon/user-agent-string-switcher/versions/

https://github.com/e3kskoy7wqk/Chromium-for-windows-7-REWORK/issues/65

@GH0st So, here is a workaround. Create a SSUAO for cloudflare.com, i.e. general.useragent.override.cloudflare.com with the following user agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:68.0) Gecko/20100101 Firefox/68.0 Test it and report here! And if that doesn’t help, try disabling uBO completely, i.e. the extension in the Add-ons Manager, if you’re using it. An uBO set too aggressively made it impossible for me to complete the CF challenges.

Don’t be misled! Quoting from old, already closed and resolved at the time issues serves no purpose and is misleading. I’ve opened a new issue on GitHub today: https://github.com/Feodor2/Mypal68/issues/723 One thing is clear. If you use a browser that doesn’t support Cloudflare’s security nonsense, you won’t get very far in these days, and it’ll be useless. Indeed, Cloudflare = Cauldron Filthcare.

New build of BOC/UXP for XP! Test binary: MailNews Win32 https://o.rthost.win/boc-uxp/mailnews.win32-20260725-40a79c75-uxp-30427c0213-xpmod.7z BNavigator Win32 https://o.rthost.win/boc-uxp/bnavigator.win32-20260725-40a79c75-uxp-30427c0213-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.win32-20260725-id-656ea98-uxp-30427c0213-xpmod.7z IceApe-UXP(suite) https://o.rthost.win/hbl-uxp/iceape.win32-20260725-id-656ea98-ia-c642e3c-uxp-30427c0213-xpmod.7z source repo (excluding UXP): https://github.com/roytam1/icedove-uxp/tree/winbuild https://github.com/roytam1/iceape-uxp/tree/winbuild

@Dave-H If you absolutely want to stick with Panda – presumably because you still have a valid licence – then I’d uninstall it completely, give everything a thorough clean, then reinstall it, but without allowing any connection to the internet. Then set the cache folder to read-only for everyone, especially for the System account, or if that doesn’t work, delete it completely. Then set all your exclusions and only then connect to the internet. No updates must be carried out locally. Perhaps it will then run stably. But to be honest, this isn’t of course a good workaround, as it unfortunately results in a loss of functionality. Personally, I’m done with Panda and would only go back if an official fix has been released or the update server has been corrected.

This absolutely confirms what I’ve already said here and what many other customers have reported across the web. At first glance, Panda Support seems great, but on closer inspection it’s a joke. For home users, it’s practically non-existent. They probably focus more on enterprise customers. Money makes the world go round.

@VistaLover I reversed this commit which got partial functionality but GCC 16 is still problematic for 32-bit builds. Here is a smaller build using GCC 15 for you to try. https://files.videohelp.com/u/69760/ffmpeg-8.2-2278-a09be9b91e-static-VISTAx86.7z

There are new Vista builds linked in the previous post. If you can let me know if ffplay is working for you.

... 12.5 is for XP no doubt. However I already threw it off my box because it's too cut-down. This statement is not really true. Although IOBit Malware Fighter Pro 12.5 can still be installed strangely enough under Windows XP, it says that this version is not compatible with Windows XP, if the user tries to enable real-time malware protection with IObit & Bitdefender engines or advanced protection against ransomware. It also advises to install IOBit Malware Fighter Pro 11.4 as the last XP-compatible version. And so, after all, this is what applies:

@VistaLover Thank you for reporting the problem. I will investigate and make new builds.

FYI, today I contacted Panda Security by telephone via AnyTech365 and informed them of the problems with Panda Dome's PSANHost process on Windows XP, which have been ongoing since 2 July. There are indeed free of charge helpline numbers for all sorts of countries, and as many users as possible should get in touch. Here is the page with the telephone numbers: https://www.pandasecurity.com/en/homeusers/free-antivirus/support/ So please leave a message there. Call them! The more people who get in touch, the better.

I've now tried as a test removing the cache folder completely from my Panda installation. Rather unexpectedly, the program has not recreated it. It's been running for ages now like that with no crashes, but it's not updating, presumably because it has nowhere to put the files! Doing a manual update from the option in the settings does nothing. I guess I could leave it like that, but of course, like the version of Malwarebytes I used on XP for years, it will eventually become useless for protection because of no updates. I tried Avast on my system a few years ago and never got it to work on my desktop system. Even the supposedly compatible version for XP would not run, its core service obstinately refused to start, reason unknown. I even worked with Avast's technical support people about it, and even they couldn't determine what the problem was. I tried a test install on my netbook under XP, and that worked perfectly, so what the problem was with the desktop remains a mystery.

Here are a few words about my plan C. More than 3 years ago, my good old friend @D.Draker had recommended the anti-malware programme Vir.IT eXplorer. I still had it on my internal list, but as there were no further tests or information available, I couldn’t report on it. At that time, it also seemed to have some temporary issues with VB100 certification, but these had long since been resolved. I was also worried that support for Windows XP had long since been discontinued, as it is no longer listed as an OS on the website. So I downloaded the very latest version 9.5.981.0 and was delighted to discover that the programme is still compatible with Windows XP. For some days, I have been testing Vir.IT eXplorer Lite. More on this coming soon.

@Tangy Perhaps you can try Panda Dome 21.01.00. I am under the impression that this version is working better on Windows XP. It's definitely worth a try.

BTW, the cache folder inside the Panda programme directory is not the only destination folder for the updates. Files are, of course, also stored in various subfolders under Documents and Settings\All Users\Application Data\Panda Security.

Not to get anyone's hopes up, but I may have found a fix, or at least a workaround. In the Panda program files folder, there is a subfolder called 'cache'. I looked at that and thought, 'clearing out caches is often a good idea'. After first backing it up, of course, I deleted all the files in it. The program then ran on a reboot, with no crashes. All the instabilities seemed to have gone away. The only thing that happened was that Panda started prompting me to OK the running of a lot of program files, so maybe that's where that data was stored before. After OKing all the prompts, things were still good, but surprisingly the cache folder still didn't seem to have anything in it. The folder originally contained a .mngr file, a .util file, a .code file, and several sets of .act and .sig files, which were in pairs with the same filename. The first three files were dated from the date of the program's installation; of the other pairs, one was from back in February, and the others were all from July 3rd this year. I put all the files back, except the recent ones, which I was suspicious of, and everything is still fine. So, was there a corrupted file or files in there?