AnX

Has anyone tried using one of @roytam1's browsers and spoofing the user agent?

Probably. Windows 9x seems to act up on the system in all sorts of weird ways. Even though I can install ME, trying to open up an MS-DOS command prompt hard locks the system, and therefore anything that relies on the DOS prompt doesn't work. I'd also need the SATA/AHCI driver to get proper disk access. For now, I'm only going to be running XP on that box until I can purchase the needed software.

New version of Basilisk running excellent. As always, thank you @roytam1. Testing on Windows XP SP3(4) (AHCI) installed to the SSD of machine #2 in my signature.

Okay, so I tried what you told me to do, and I found out that the system.ini file gets corrupted on its own, being replaced by a bunch of garbled junk. At this point, I'm pretty much at my wit's end with this system. Recently, two windows xp installs have failed for strange reasons, and I think this machine has some bigger issues with it that may be causing these problems. EDIT: was able to get XP on the SSD but only in AHCI mode. I think the IDE emulation on this platform might be broken.

Will do, thanks. I already had the majority of devices disabled, but left a few dormant. Those could be the culprits.

I see. In that case I'll purchase it when I can, since it seems that is the only proper solution to this "madness".

UPDATE: I have been successfully able to get Windows Millennium Edition to install, using the setup /p i switch. I still don't have 32 bit file access, however I think if I can find an ATA driver, I should be able to fix it. Regardless, the OS is working perfectly (ironic, as it's supposed to be the other way around, lol). Anyway, I still want to get Windows 98 installed. Any solutions for the registry error that it keeps coming up with? The install media is 100% clean and legit, I've tested it multiples of times. Edit: there are still some strange issues with Me. DOS prompts don't work at all, they freeze the system. Also, anything in general that is trying to directly access the hardware doesn't seem to run right (like the system information tool, cpuz, etc.). That might also be related to the disk access issue.

Excellent work. Did you have any troubles getting it running?

I've been trying to install Windows 98 SE on the M5A99FX PRO R2.0, socket AM3+. I have a secondary machine that I wish to use for Windows 9x and other older Operating systems. The specs are in my signature. Unfortunately, I run into a lot of issues when I go to install Windows 98. Windows 98 is being installed on the 60GB SSD, in this case. I've tried the following approaches. -> When I install using the setup /p i switch, the OS seems to be installing normally with very little to no problems. But then, out of nowhere, SCANREG starts and claims that the registry is corrupt, and it can't fix it. The entire install is trashed for no apparent reason. -> When I install using the setup /p j switch, which forces ACPI, the OS starts detecting devices and hangs. I then reboot into Safe Mode, remove the ACPI stuff, reboot again, and this time setup can successfully install the OS (there is still a registry error, but it gets fixed this time). Windows 98 also re-enables ACPI, and it works, partially. Unfortunately, Windows 98 in this config proves to be quite unstable, where even a small change in the USB devices could hang the system. However, by far this is the setup I've had the most success with. -> When installing using setup /p a;b for safe and selective detection, I tried to selectively disable detecting certain devices which may be problematic. Unfortunately, not only did it take much more to seemingly get running, but the SCANREG error creeped up and trashed the install. In all cases, I've disabled the 32 bit file access drivers. I use the MaxPhysPage trick to limit RAM, and can confirm it works. This is at 100% stock, factory CPU settings, no overclock, and SATA is on IDE mode. The platform runs Windows 2000 and XP fine (obviously newer than those, too) What can I do to get Windows 98 to install correctly? In any non-ACPI case, it simply trashes the registry all by itself. In any ACPI case, it either haphazardly works or simply hard locks, and it takes so much effort and trial and error to get it to somewhat work, too.

Follow-up: Windows 95 and Windows ME install without a hitch in VMWare. However, when you run a 16 bit program in 8086 virtual mode, you still risk it locking up the system. The only solution is to run said programs in real mode. Windows 98 first edition is extremely buggy with the VME bug, and I haven't been able to fix it at all. Also, the VME bug doesn't exist on 2nd generation Ryzen CPUs or later. It is solely a Ryzen 1000 series problem.

You should install only the latest extended kernel. Indeed, every succeeding version contains the previous version's patches.

I didn't expect fixes for Windows 7 to ever happen, but here we go. The Ryzen APUs, atleast on MSI motherboards, are now Windows 7 compatible. Unfortunately, the integrated graphics driver is not compatible, which kind of negates the whole point of these CPUs. Still, that means that this is merely a bug, and not a lockout, which makes me optimistic for future Ryzen CPUs.

I'd like to know if there's still a modern browser or a fork of a modern browser that can support Windows 2000 without the kernel extensions. @roytam1's builds for XP run great under 2000 + ExtKern, but there hasn't been an updated build for 2000 w/o ExtKern in particular for a while.

You must enable the ACPI Multiprocessor Hal in Device Manger. It's under the "Computer" section. For you, it'll likely be ACPI Uniprocessor. All you have to do is update the driver and select the ACPI multiprocessor Hal, which should give you full multi core support, without having to do anything else.

Don't mention it. This'll probably be useful for anyone trying to run 9x bare metal, too. I believe Ryzen 2 and TR/EPYC don't have the VME bug, or it's much less pronounced. But Ryzen 1.0 does, and the fixes don't completely mitigate it.

The AMD Ryzen CPUs have a VME bug, which was mostly fixed in the latest revisions and BIOS updates, but hasn't been entirely fixed for hybrid 16/32 bit OSes (such as Windows 9x). A solution to this is to disable the 32 bit protected disk drivers (go into Safe Mode, then System Properties > Performance > File System > Troubleshooting > Disable all 32-bit protected mode disk drivers), and then install the 48-bit LBA fix, as well as all of the previously released Windows 98 updates. Then, you should be able to re-enable the 32 bit drivers. This may not guarantee 100% stability, but it seems to work fine for me.

I'm looking at getting my Ryzen system working with Vista. Getting the USB 3.x drivers in is probably gonna be the biggest hurdle. It runs 7 like a charm, but for a friend of mine who likes Vista, I'm looking to test Vista on it too. AMD uses ASMedia for a lot of their chipset, and ASMedia is known to support Vista to an extent, so we could get lucky.

I have a feeling this would be more appropriate on the XP forum, but regardless, it had to do with it popping up some strange storage related errors. Even updating AHCI drivers did nothing. I'll install XP again as a secondary OS and see if I can get it working the next time around.

I was starting to like Windows 10. But some caveats have caused me to go back to Windows 7. 1. For some reason, when I go to open my folder with all my documents, Explorer would crash. 2. Windows 10's updating system was horrible. Because my Internet is limited, I need more control on how Windows 10 downloads its updates. But when I go to do that, it decides to just sit there and use up 30% of my CPU doing nothing. 3. Who knows how much data was being siphoned out... however I haven't really recorded any significant signs of that happening. 4. Compatibility with older programs is quite subpar. Anyway, that's just me. I'm pretty sure Windows 10 has been working for many others, but I'm gonna wait longer. 'Till then, 7 is a solid OS.

Thanks for the heads up, BWC. It's really a shame how MS let Windows 2000 to die off. The fact that they claimed 2000 had an issue which was fixed just so they could sell more copies of server 2003 is downright misleading. Furthermore, I'm relived that Windows 2000 supports high cache CPUs, as I run it on an FX-8320 box because of my 26 year old compiler that hates modern OSes (and virtual machines). I would run XP, but XP simply produced issues of all sorts on that machine for whatever reason. Also, I must clarify that the BSOD only happened while installing a sketchy HPET driver. My simple solution is to simply disable the device altogether, and Windows 2000 runs solid as a rock. Additionally, my AFD.SYS is version 7204, the fixed version.

You'll be able to run Windows 10 x64 on your machine just fine. http://www.geforce.com/drivers/results/112596 These are the drivers needed for your graphics unit.

The issue with Win2k's MPS HAL is well documented, and typically happens in the following scenarios: -Certain BIOSes (such as "insydeH2O") -Conflicts with the Serial/Parallel port (this was documented by bluebolt and could actually fix your problem, if you have a way to disable those ports) It's unknown why Win2k (and NT4)'s MPS kernels do this, however you gotta understand that they were built for multi processor systems years ago. Perhaps a change in how Windows XP recognizes processors is what fixed the issue for that OS. Nonetheless, Win2k does end up supporting most MPS/Multi-core systems, provided that they are either desktops (custom built ones in the case of modern units), or laptops prior to 2008 (some 2011-2012 ones support Win2k, but it varies). Even though it's probably too late to ask, I'd love to see a fix up of the Win2k MPS HAL to repair these issues, and possibly even bring in proper HyperThreading support. It's a long shot and likely not even worth it for this OS, but it would be interesting to see the result.

I was browsing the web, and came across an article which points to a flaw under Win2k that could cause it to BSOD with the 1E error "under a processor that is very fast or with a large L3 cache". https://support.microsoft.com/en-us/kb/924439 So, I was wondering if anyone over here has actually experienced this anomaly with Win2k. I will admit that I got some KMODE_EXCEPTION_NOT_HANDLED errors on my FX-8320 when I tried to install an HPET driver, but without it, Win2k runs solid there.