All Activity

well technical speaking actually it would be possible to make a check what technology in the XMM registers are possible there was recently a problem like that with a microsoft file (in the supermium posts) where actually SSE4.1 instructions are used more professional programms (and useally you would expected something like that from a very big money concern like microsoft) check what instructions are available (the x265 encoder/decoder does so for example) after that then the right routine is called up often it is a compiler setting problem one method to avoid this one would be to create an .obj file (what contains the functions without SSE/MMX ect.) then at first you make the check if SSE can be done if not you call up that .obj file´s function if you have SSE you call up your normal function thus you can avoid creating multiple complies like having a SSE version and a non SSE version

Bug Report: There seems to be a bug in the reimplementation of the File Explorer Menu Bar in the past several versions of SAB. If you are trying to paste a file in File Explorer and you pull down the Edit menu on the Menu Bar, the Paste option will be graded out. But if you right click in File Explorer, on the context menu the Paste option is highlighted and works normally. Pasting from the Menu Bar, Edit menu is very convenient and it would be great to have this bug resolved.

I tested your audio driver on WindowsMe installed on real hardware with Realtek ALC662 HDA on-board. The driver is installed without any problems as "WDM Sample Driver for HD Audio", the system boots and works fine but there is no sound in the speakers (tried all the audio jacks). Anyway, thank you very much for your contribution to the win9x community

My old, underpowered Windows XP computer with a Pentium 4 single-core processor is a multiboot machine. Five OSes were independently installed and can be booted totally indepedently by the best bootmanager I have been using from the very first: BOOT-US. The OSes are Windows ME, Linux (XUBUNTU and Q4OS) and Windows XP twice. No problems here with starting OSes in the first place. BTW, my Windows XP computer is now 25 years old and fully working. I don't believe that modern computers will reach such an age.

FWIW, and off-topic, IMO any operating system version which is no longer supported could be said to be an 'older operating system'. That now even includes early versions of Windows 11!

I should add, we all should consider that Win10 from 2016 through 2019 shoud all be reconsidered here at MSFN as *QUALIFYING* for "older OS's". Here at MSFN, you say "Win10" and all the XP and Vista folks seem to outcast you as not using an "older OS". I am not a fan of any of the "H" versions of Win10+. ie, anything that has a version numbering like 21H2, 22H1, 22H2, 24H2, 25H2, et cetera. But Win10's that PRE-DATE that numbering scheme is VERY MUCH SO an "older OS".

Does not appear to be the same issue. For me, the issue was *ONLY* on Supermium's SETTINGS PAGES and not on visited web sites. Though it is possible that it is related somehow - a font issus is a font issue, whether it be an external page (visited web) or an internal page (settings/flags). Truth be told, I *rarely* actually visit any web pages using Supermium (it did not play well with my old hardware and there is NO USE to use it on NEWER hardware). As it's "last straw", I upgraded it from Win XP x86 to Win 10 *LTSB 2016* and that did MUCH BETTER than trying ported browsers trying to prolong XP. Will be *weeks* (if at all) before I can repurpose my old Intel Core2 Quad Q6700 where this font-indent issue exists. BUT... personal experience being my guide, it will very likely STAY on Win 10 *LTSB 2016* and it will never again "waste my time" attempting to prolong XP. "Been there, done that", as the saying goes.

Perhaps the issue is that FF 39 does not support TLS 1.3, so websites that use this protocol cannot be opened. FF 2 has an encryption engine update, so websites with TLS 1.3 can be opened there. To enable the browser to use system certificates, find (or create) the security.enterprise_roots.enabled parameter in about:config and set it to true.

If you mean this: https://github.com/win32ss/supermium/issues/1596 This has been fixed in 138 R6. The problem was in the instructions; if the processor did not support SSE4.2, this bug could occur.

my UXP IA32 builds works on any processor that has CMOVcc, which Pentium II and Cyrix 6x86MX/MII and later will work. and in the past I did tried to make firefox compiling on VC2013 as far as possible: https://github.com/roytam1/fx-vc2013 and NM27 also has many changes in fx48 and even later versions (thanks arctic-fox forks for the job). but too many efforts put in and no much outcome achieved so I focused on UXP again.

yes it is always an issue because GetLogicalProcessorInformation is missing in XP SP2. you may either update kernel32.dll with that API available(i.e. WindowsXP-KB936235-v2): http://kmeleonbrowser.org/forum/read.php?19,143253,page=10 or patching your target application with kernelxp and patchk32. https://github.com/roytam1/rtoss/tree/master/kernelxp a patched SP52 running on just-installed XP SP2 VM: or unpatched SP52 build on WindowsXP-KB936235-v2-x86-ENU installed XP SP2 VM:

Speaking about forks,Yansu has been made,and it's a fork of the last version for non-SSE2 PCs,Firefox 48. This might be irrelevant due to 45esr getting forked and Serpent IA32 exists but 48.0 seems forgotten by developers...even those who forks... Maybe @roytam1 can fork it and add his own changes to it.

I'm using SP2... maybe that's the issue?

Normally Readyboot should help with slightly faster boot times and faster recovery from hibernation by pre-fetching files. But ironically it can cause the exact opposite if too much logging is happening, so in your case you are not losing anything. This feature might have been useful to some extent when HDDs where the norm, but with SSDs or even NVMe it shouldn't make much of a difference.

Thanks for the hint! But I don't think the issue has already been fixed. However, one should definitely take a close look at the new "Legacy Installer", if available. Ok. I checked the "Legacy Installer" FREEAV_LGC.exe but it wasn't updated. Only the FREEAV.exe installer.

I occasionally run XP x64 as a dual-boot on my 13yr old i7-4770 machine. It used to run XP x64 exclusively but has since been upgraded to an OLDER version of Win10. But the "problem" I ran into (here at MSFN) during those exclusive-use XP days on that i7-4770 "era-correct" computer (XP went EOL *after* my computer manufactured date) was that I somehow didn't "qualify" to be running XP in the first place. Or that VM's somehow also don't "qualify" (regardless of host being XP or not).

I would also be interested in reports from people who use Windows XP on more powerful computers. As always, my results come from my old, underpowered Windows XP computer, which simply refuses to leave me. But this computer is perfect for diagnosing such errors that you might not notice on more powerful computers. The offline installer can't be found on Panda's homepage. Here is a website where you can get the old offline installer of version 21.01.00: https://panda-free-antivirus.sooftware.com/windows/download/338378 You have to enable the checkbox "Offline installation" during installation. After finishing installation, the programme won't update to a more recent version, even when the automatic update function of the programme under settings is in use. Thanks for the hint! But I don't think the issue has already been fixed. However, one should definitely take a close look at the new "Legacy Installer", if available.

I suppose your investigation results are quite valuable to XP SP3 users on old and under-resourced H/W; so, thanks for your time spent ; however, unless one had downloaded to disk v21.01.00 when it was still current, it's now a real task to find working/valid links to actual (off-line) installers to it ; I did a simple G-search for it and ALL results on pages 1-3 point to (mostly) stub or off-line installer-links hosted on the vendor's (PANDA) servers; but these tend to download the latest version of the software, not the older one we're after ; I did succeed in the end in finding links to offline installers for v21.01.00, but these were self-hosted on Iranian and/or Russian sites, so, as you can understand, I won't share them publicly here ; other than that, is there a special setting one should configure in v21.01.00 so that it doesn't auto-update to a more recent version? FWIW, v23.00.00 has now become the latest (digitally signed 22/12/2025) ... Kindest greetings.

I can't tell whats wrong on your side but it is working here.

Hi roytam. Merry Christmas to you and everyone here. Installed this latest version of Basilisk but it didn't start. Had to go back to the previous one.

I'm sorry, somehow I thought I'd just normally decompress them instead of installing them. :-) Since MilkChan has built those fine update packages, I'd say SSE (b5107) for slower computers and SSE2 (b6912) for faster. Could you please test both and see which is faster. I think b5107 might be faster (it is also the last build in non-ESU fully updated Server 2003 x86).

Thanks for the list ... I downloaded 5 files, from newer to older and will see if any actually install. I hope one download will install, if not I will download more from the list to find one that will install. No, I just wanted to install OpenSSL on my XP setup.

As you know, the speed of ScanDisk under Windows 9x depends on the intensity of mouse movement. I made a short recording demonstrating this effect: As you can see in the recording, the speed of the scan increases significantly when you move the mouse. The question is, is it possible to fix this behavior so that the scan always runs at maximum speed and is not affected by mouse movement? Thank you.

Merry Christmas and please enjoy a freshly specially experimental build!

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-20251225-e8bf1a87f-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk55-win64-git-20251225-e8bf1a87f-xpmod.7z repo: https://github.com/roytam1/basilisk55 Repo changes: - ported from UXP: Issue #2306: Implement WeakRef functionality (5eab8c0a) (6593da0bd) - ported from UXP: Issue #2306: Add pref-gated WeakRef with safe stub when disabled (5895253f) (b07afbb3e) - import from UXP: Issue #2306: Fix spacing (b9ded9c3) (d68445246) - ported from UXP: Issue #2306: Remove unneeded addition of streams_ in ContextOptions() (d43e57f8) (df4c765a7) - import from UXP: Issue #2306: Add test js (c171ddea) (223b2a470) - ported from UXP: Issue #2863 - Part 1 - Update patches for ICU 78.1 (53662b13) (3a4140453) - import from UXP: Issue #2863 - Part 2 - Filter ICU data for build (d5bcbfd0) (67724cfb9) - ported from UXP: Issue #2863 - Part 3 - Patch in non-dll suppression (fe4deee3) (3fada1a0e) - import from UXP: Issue #2051 - Part 3 - Use pre-built big-endian ICU data file for big-endian builds. https://bugzilla.mozilla.org/show_bug.cgi?id=1264836 Disable unused ICU code during build. (7299df47) (0a7458859) - import from UXP: Issue #2863 - Part 4 - Use recommended ICU build options flags. (759d7984) (7b5f67620) - import from UXP: Issue #2863 - Part 5 - Update reading source objects (35cb857c) (452b32a22) - import from UXP: Issue #2863 - Part 6 - Make ICU with multiple jobs for faster rebuilds. (72f61359) (b5331bdb6) - import from UXP: Issue #2863 - Part 7 - Don't delete files now that they're filtered (c8491cff) (8d43b8b12) - import from UXP: Issue #2863 - Part 8 - Add note to run `make_intl_data.py langtags` (5a3d1021) (d15c0dc7d) - import from UXP: Issue #2863 - Part 9 - Retain util_props.cpp (a8a1c5c2) (ca5715a00) - import from UXP: Issue #2863 - Part 10 - Add New Line Breaker Entries (afa02844) (8c4575973) - import from UXP: Issue #2863 - Part 11 - Remove no longer needed locale maximization when removing likely subtags. (b31a7165) (d59467deb) - import from UXP: Issue #2863 - Part 12 - Filter SCRIPT_CODE extensions from CLDR (25caba3d) (60e07b23a) - import from UXP: Issue #2863 - Part 13 - Remove non-IANA Zones (77f1baf2) (631486a72) - import from UXP: Issue #2863 - Part 14 - Update switch to handle new UNumberFormatFields entries. (f47df7a6) (0b6e69a35) - import from UXP: Issue #2863 - Part 15 - Update special casing script (98fc5db1) (6d0cdf00f) - import from UXP: Issue #2863 - Part 16 - Update locale canonicalization algorithm for changes from CLDR 39 (3f06fa4c) (fb4a3c5c9) - import from UXP: Issue #2281 - Remove the use of non-standard namespace std::tr1 in Google's WebRTC gtest code (cf85580d) (c03434692) - ported from `custom` branch of UXP: Issue #2863 - Part 17 - Update ICU source to 78.1 (a29a8049) and fixes (9f672787) (38bdf911b) - gmp: stop unified build to avoid build bustage (757d587fe) - import from UXP: Issue #2863 - Part 18 - Update Unicode in Spidermonkey (852e6904) (6a04657eb) - import from UXP: Issue #2863 - Part 19 - Update CLDR to 48 (666b590b) (89b712a31) - import from UXP: Issue #2863 - Part 20 - Update special casing in Gecko (ec13c274) (f97d9374f) - import from UXP: Issue #2863 - Part 21 - Update Unicode property data in Gecko (42d7950b) (1fabdf19c) - ported from UXP: Issue #2863 - Part 22 - Update tzdata source to 2025c (929b21c7) (d1aaae52d) - ported from UXP: Issue #2863 - Part 23 - Update tzdata in Spidermonkey (f7284978) (ec0aa8a35) - import from UXP: Issue #2863 - Part 24 - Remove special-casing for Etc/UCT (c77fea87) (8098de289) - ported from UXP: emoji-17.0 - Update Twemoji font to unicode 17.0.0 (36a295e4) (8549af4bc) - import from UXP: Issue #2857 - Implement `inset-block` and `inset-inline` CSS shorthand (5a832a6e) (7e09c7b47) - dom/webrtc: explicitly exporting `NormalizedConstraintSet::Range<bool>::Merge()` to fix build (e8bf1a87f)