BetaArchive

I'm trying to install Memphis Build 1400 to run some tests. SETUP completes properly, but every time I try to boot to the Desktop, EXPLORER crashes with an Illegal Operation error before the Desktop can be loaded. This even happens in Safe Mode.

Anyone know how to prevent or circumvent this? I'm installing it on real hardware, and it has to be on real hardware for this test, so VM-specific solutions are not relevant.

I have tried using EXPLORER.EXE from Build 1411 but the same crash still happens.

Hm. I installed this build on my PII and I had no such issue. Is your install media corrupt?

Wheatley wrote:Hm. I installed this build on my PII and I had no such issue. Is your install media corrupt?

Shouldn't be, I downloaded it from here and copied all of the files to the HDD for the installation. This happened before several months ago on a previous attempt to install it as well, but then it wasn't essential to have this exact build loaded so I moved on to another one.

Interesting that it works on a PII - the hardware I'm using is quite a bit newer than that, but builds 1387 and 1411 run on it without problems, so 1400 should as well...

What are the specs of the PC? And did you change the BIOS date? Sometimes EXPLORER will not open if the bios date is not set.

Keep in mind that early builds like this can be picky about the hardware you run them on. I had this exact same problem with build 1400 in Virtual PC 2007. Never tried it on a real PC before so I don't know what's going on.

giantsteen wrote:What are the specs of the PC? And did you change the BIOS date? Sometimes EXPLORER will not open if the bios date is not set.

I changed the BIOS date prior to installing, and the timebomb is no longer an issue. I also have a DATE command called in AUTOEXEC.BAT that resets the date back on each boot in case of other issues. This doesn't seem to have any effect on the EXPLORER crash.

The specs of the system are as follows:
MSI K8N Neo2 Platinum Motherboard
AMD Athlon 64 X2 4800+
512MB DDR RAM

Now before anyone complains about that, as I said earlier I am able to run Builds 1387 and 1411 on this machine without problems, using only RLoew's CPU speed patches. In fact the screenshot from Build 1351 I posted in that thread is on this same machine.

I do realize though that this particular build could have some issue with this setup that the other builds don't, so I'm hoping that someone else will have already encountered this and have some idea what might fix it.

TheWindowsXPGuy wrote:Keep in mind that early builds like this can be picky about the hardware you run them on. I had this exact same problem with build 1400 in Virtual PC 2007. Never tried it on a real PC before so I don't know what's going on.

What were the specs of the VM you created? How much RAM did you assign to it?

LoneCrusader wrote:

TheWindowsXPGuy wrote:Keep in mind that early builds like this can be picky about the hardware you run them on. I had this exact same problem with build 1400 in Virtual PC 2007. Never tried it on a real PC before so I don't know what's going on.

What were the specs of the VM you created? How much RAM did you assign to it?

I can't remember for sure, but it was around 16-32 MB.

TheWindowsXPGuy wrote: I can't remember for sure, but it was around 16-32 MB.

Hmmm... Well that at least should eliminate RAM as a potential issue in this case if yours was set that low and you still encountered the problem.

I have same problem and I couldn't find a solution. I'll try VMware and write the result here.

Probably is a problem with RAM memory (512 MB is very big for 98 betas) so, remove 512 MB from computer and put 64 or 128 MB of RAM. Probably, this is a problem.

Maxxyz wrote:Probably is a problem with RAM memory (512 MB is very big for 98 betas) so, remove 512 MB from computer and put 64 or 128 MB of RAM. Probably, this is a problem.

No, probably it is not. 512MB of RAM is usable for Windows 95 and any build thereafter. Only this one build suffers any problem on this machine, and I have installed EVERY OTHER Memphis build existing here at BA on this machine without problems. Also, if you read above, the same problem was present on a VM with even less RAM than you advise.

I do plan to attempt reducing the RAM if no other solution is presented, although I do not expect it to change anything.

Do you have any optional hardware in your system? (eg. soundcard, usb card,...) Remove as much as possible and then try again. Maybe this can be a problem.

I am having no problems whatsoever

Jensvde wrote:Do you have any optional hardware in your system? (eg. soundcard, usb card,...) Remove as much as possible and then try again. Maybe this can be a problem.

I have no cards other than the Video Card. All onboard devices on the motherboard such as Sound, LAN, 1394, SATA, etc, are disabled in the BIOS except USB. I need USB for the experiment I'm doing.

zacb27 wrote:I am having no problems whatsoever

Obviously you're running it on a VM. 1400 must run on real hardware for the experiment I'm doing. A VM installation would be useless, unless something in the specs of the VM you set up might lend a clue.

zacb27 wrote:I am having no problems whatsoever

What site did you use for downloading Memphis 1400?

I assume he either a) had it previously or b) downloaded it from Winworldpc, as he didnt have ftp access when he took that screenshot.

Maxxyz wrote:Probably is a problem with RAM memory (512 MB is very big for 98 betas) so, remove 512 MB from computer and put 64 or 128 MB of RAM. Probably, this is a problem.

Try reinstall build.

cantasan99 wrote:I have same problem and I couldn't find a solution. I'll try VMware and write the result here.

He didn't try to install it on VM, he tried to install on real PC.

Also, LoneCrusader, did you try disable ACPI?

BettaFish1902 wrote:

cantasan99 wrote:I have same problem and I couldn't find a solution. I'll try VMware and write the result here.

He didn't try to install it on VM, he tried to install on real PC.

Also, LoneCrusader, did you try disable ACPI?

Yes, I tried disabling ACPI by INF-editing based on my guide in the "other" 1400 thread. It doesn't seem to have any effect on the EXPLORER crash.

Also of note - I did try another experiment, this time installing 1400 to a VM (VMware Workstation 7) to see if the same crash would occur. The VM was hosted on a 3.4GHz P4 machine, and I assigned 512MB of RAM to it. I did not apply any OS-specific tweaks and set the VM up as if it were a Windows ME machine rather than 95/98 to be certain no "workarounds" would be applied. The EXPLORER crash does not happen at all on the VM, so CPU speed and RAM are apparently not related to the problem.

LoneCrusader wrote:

BettaFish1902 wrote:

cantasan99 wrote:I have same problem and I couldn't find a solution. I'll try VMware and write the result here.

He didn't try to install it on VM, he tried to install on real PC.

Also, LoneCrusader, did you try disable ACPI?

Yes, I tried disabling ACPI by INF-editing based on my guide in the "other" 1400 thread. It doesn't seem to have any effect on the EXPLORER crash.

Also of note - I did try another experiment, this time installing 1400 to a VM (VMware Workstation 7) to see if the same crash would occur. The VM was hosted on a 3.4GHz P4 machine, and I assigned 512MB of RAM to it. I did not apply any OS-specific tweaks and set the VM up as if it were a Windows ME machine rather than 95/98 to be certain no "workarounds" would be applied. The EXPLORER crash does not happen at all on the VM, so CPU speed and RAM are apparently not related to the problem.

Did you try comparing the settings used for the two virtual machines (the one with the "Windows Me" option set vs. the one with the "Windows 98" option set)? Perhaps if we knew what settings the virtual machine software was changing to allow the operating system to work, then we might be able to discover the root of the problem.

LoneCrusader wrote: Yes, I tried disabling ACPI by INF-editing based on my guide in the "other" 1400 thread. It doesn't seem to have any effect on the EXPLORER crash.

Also of note - I did try another experiment, this time installing 1400 to a VM (VMware Workstation 7) to see if the same crash would occur. The VM was hosted on a 3.4GHz P4 machine, and I assigned 512MB of RAM to it. I did not apply any OS-specific tweaks and set the VM up as if it were a Windows ME machine rather than 95/98 to be certain no "workarounds" would be applied. The EXPLORER crash does not happen at all on the VM, so CPU speed and RAM are apparently not related to the problem.

Then, did you try re-installing build, as Maxxyz mentioned before?

WinPC wrote:Did you try comparing the settings used for the two virtual machines (the one with the "Windows Me" option set vs. the one with the "Windows 98" option set)? Perhaps if we knew what settings the virtual machine software was changing to allow the operating system to work, then we might be able to discover the root of the problem.

I'm not using two VMs, I'm using one VM and one real hardware. The only reason I chose to set up the VM as "ME" was to ensure that no CPU speed workarounds would be applied to the VM. I think? some VM software does this for 95, maybe 98? I don't know what differences there are between 98 and ME VM settings, or if it would even be relevant here. Comparing the ME VM settings that DO work against say, settings for an XP VM (IF the "XP VM setup" did NOT work, which I have no idea whether it would or not) might be relevant.

BettaFish1902 wrote: Then, did you try re-installing build, as Maxxyz mentioned before?

Yes. Reinstalling just leads back to the same problem.

Windows 95, then Memphis 1400?

LoneCrusader wrote:

WinPC wrote:Did you try comparing the settings used for the two virtual machines (the one with the "Windows Me" option set vs. the one with the "Windows 98" option set)? Perhaps if we knew what settings the virtual machine software was changing to allow the operating system to work, then we might be able to discover the root of the problem.

I'm not using two VMs, I'm using one VM and one real hardware. The only reason I chose to set up the VM as "ME" was to ensure that no CPU speed workarounds would be applied to the VM. I think? some VM software does this for 95, maybe 98? I don't know what differences there are between 98 and ME VM settings, or if it would even be relevant here. Comparing the ME VM settings that DO work against say, settings for an XP VM (IF the "XP VM setup" did NOT work, which I have no idea whether it would or not) might be relevant.

BettaFish1902 wrote: Then, did you try re-installing build, as Maxxyz mentioned before?

Yes. Reinstalling just leads back to the same problem.

Are you sure that the problem isn't with the drivers? Sometimes, driver issues can create problems when installing or using any version of Windows 9x, especially on a real machine. I myself had a Windows 95 installation become unstable simply because of a video driver that was only partially compatible with the system's video adaptor.

BettaFish1902 wrote:Windows 95, then Memphis 1400?

I don't do "Upgrade" installs unless there is no other option.

WinPC wrote:Are you sure that the problem isn't with the drivers? Sometimes, driver issues can create problems when installing or using any version of Windows 9x, especially on a real machine. I myself had a Windows 95 installation become unstable simply because of a video driver that was only partially compatible with the system's video adaptor.

No drivers have been installed other than those that come integrated into Memphis 1400. I can't boot to the Desktop in Normal or Safe Mode, so I haven't even tried to install anything yet.



Every other RTM or Beta Windows 9x version from 95C OSR2.5 to Windows ME runs fine on the machine in question. Only Build 1400 has some kind of problem, so it leads me to believe some file has a bug. The fact that using EXPLORER from 1411, 1415, 1423, 1387, 95 OSR2, and a 95 OSR2 HotFix do not solve the problem seems to indicate that the problem does not lie in EXPLORER itself, despite the fact that it is the file reported in the error.