LoneCrusader

When any devices lacking drivers are detected during 9x installation, they are added to the Device Manager as "Unknown devices." You will not be prompted to install any drivers for them. To force redetection and get a prompt for drivers, you will need to manually remove any "Unknown devices" from the Device Manager and reboot. You don't have to physically remove it each time for this. Also you could verify which .INF file is being called when the drivers are detected on the CD, and then copy it along with any other files listed inside of it to the HDD or a Floppy or whatever you prefer. But, you will still need to take the steps I gave above.

The only "case" that needs to be made is the preference of the user. You don't know why the OP wants do do it... I don't know why you would care if he does?

According to the CPUID Version History page, a 1.57 exists for 98. Versions after this have no 9x listings or comments... I clicked on some of the other builds linked on your link at OldVersion.com. Version 1.32.0 is listed working on 95 there. Version 1.33 is listed as NOT working on 95. So, given the only build between these is 1.32.1, and I have used that one on 95 myself, it would seem to support that version being the last one working on 95. Further experiments should be made, but 1.32.1 is a good place to start and should work for identifying the OP's machine.

BAH, Double Post. I'm seeing a lot of errors with MSFN tonight. SQL server errors, "Driver" errors... and sometimes the Font on the page changes when I refresh it. What's going on? @dencorso Please delete this or move it to this thread. EDIT: This error is displayed at the bottom of the page when I get the "changed Font" refreshes.

I'm not certain those versions work under 95. I haven't tested them (or if I did I don't remember) but I used v1.32.1 in this post and I believe I was using the last 95-compatible version at that time...

erpdude8's Unofficial Service Pack is incompatible with my FIX95CPU update. Installing it on a FIX95CPU system will render the system unbootable. You will have to choose which update you want or need most. If you need FIX95CPU, then you can't use the USP as well. If you don't need FIX95CPU, then you can opt for the USP instead. The Microsoft USB Support updates have been superseded by this. XUSBSUPP.ZIP - 901.0 Kb This package installs every official updated file pertaining to USB under 95. Do NOT mix it with the original Microsoft ones. (And probably should NOT mix it with the USP either... ) Firefox - first be sure to install WINSOCK2 and any other update packages you wish to install. Then: Firefox 1.5.0.12 runs natively on Windows 95. Firefox 2.0.0.20 can be made to run on 95 by doing the following: Download and run the 2.0.0.20 installer. Select custom install. Choose NOT to install the DOM Inspector and the Quality Feedback Agent. When Setup completes, do NOT launch Firefox. Find the following files and DELETE them: "nsSearchService.js" and "nsSafebrowsingApplication.js" Now run Firefox. (These instructions are taken from this original thread, but the OP there takes many unnecessary steps.) EDIT: See this also. YouTube with Flash 7/Win95

Welcome to MSFN!

The Rage 128 should have complete 95 support. The errors you are seeing with the nVidia card MAY only be the "nVidia Control Panel" that runs in the System Tray. You do not NEED this in order for the driver to work properly. You may be able to simply disable it with MSCONFIG as suggested, and get rid of the error. DX8 is DirectX 8. It is an update for Video/Audio "effects" under Windows. Usually required by games, etc. Version 8.0a is the last version that supports Windows 95. You should install it before attempting to use ANY "newer" video card such as your FX5200 because many of the newer cards require this to be updated. Download and install these: DCOM95 (required for DX8) and then DX8.

I'm assuming this issue is still limited to the AMD machine? Or does it happen on the Intel machine as well? I use VMware Workstation 7.0 on a Intel P4 system and I have never encountered any problems using 95C...

No, a .CPL is not completely "identical" to a .DLL. And in any case, the problem here is that we do not know WHAT to fix, even if we had a tool to fix it.

If the driver itself is working, you're in good shape. Try what SomeGuy and cyberformer suggested: Remember though that MSCONFIG does not exist in 95 by default. You can copy it from a 98 machine or extract it from the Q239887 Hotifx which is named 4756US8.EXE. The Microsoft links at MDGx's are dead now...

Sure it is, if you know what you need to hex edit in the first place!

This is only to run an experiment, you don't have to use or keep the resulting 95 installation. You don't even need the hardware drivers. All you need to do is find out if you can boot to the desktop after install (and installing USBSUPP) without getting the NTKERN error. If you do get it, then NTKERN may a bug with AMD that doesn't show up with Intel. If you don't get it, then you will know that it is related to the VM software.

Behavior when running under a Virtual Machine is sometimes unpredictable and is not always the same as behavior on a real machine. Can you try a native Windows 95 install on your AMD machine, just to rule out any discrepancy? I have no personal experience with AMD processors beyond the old K6-II era. I moved to P4's and have been using them since. I will ask rloew to comment on this, as he uses AMD almost exclusively...

Verify that the NTKERN.VXD being used on the AMD machine is version 4.03.1213 as contained in FIX95CPU. It SHOULD be, provided FIX95CPU is installing properly. However, NTKERN.VXD is not even used unless the USB Supplement is installed. When the USB Supplement is installed, it sometimes overwrites the NEWER version installed by FIX95CPU with 4.03.1212, which has a bug. This bug is documented in FIX95CPU's README.TXT. All that this fix solves is the frequency issue. Your own results show it works perfectly at 3000 MHz, so the issue you're having with the AMD Athlon X2 4850e is *obviously* not related to the frequency it's running at. Actually it could be, if NTKERN 1212 has overwritten the 1213 installed by FIX95CPU. This is a documented bug as I pointed out to Darth1701. However, IF this is the case, I don't see how/why it is being overwritten on one machine and not the other...

Once again, you are looking to start a fight that you know you can't win. Contradicting yourself make you look even more..................................... Sigh... Hardly a contradiction. Some files may work under certain conditions = not 100% compatible. 100% compatible = all files work under all conditions....

No, I'm not wrong. I said some NT files may work under certain conditions and not work under others. Some may work perfectly. Some may not. If an NT file works some or most of the time, that's great. You act as if I'm against using any NT files. I'm not, and I said so above. The point is: There is NO WAY TO BE 100% SURE an NT file works under ALL conditions. There are too many variables. An NT file may work fine for X program, but not for Y program. An NT file may work fine for programs A, B, C, D, and E, but not for program F. It is physically impossible to test all programs that depend on a given file. Such things require extensive testing beyond what one person can do. I know this, I'm a "project author" too. I hope that every single NT file you have in your SP works under ALL conditions. But there is no way to be certain of that. Now, you may not have seen any problems. There may not be problems. All the files you have added may work fine until another variable is added to the equation. What variable? Lets take KernelEx for example. KernelEx was designed to run on a standard Windows 98/ME installation. It expects certain conditions to exist. If those conditions have been changed by the addition of another Unofficial file, then KernelEx may not work properly. Is this your fault? NO. Is it the KernelEx developer's fault? NO. It is simply a side-effect of mixing two unofficial packages. And as usual you revert to personal attacks. Which means you are unable to debate the real issue.

The latest browsers wasn't designed for Win98, but I don't here you lecturing the Opera and Firefox users. This is why we don't see eye to eye. You always complaining about the NT files etc... They work on my system and hundreds maybe even thousands of others. Just cause you want to live in the 98 days, doesn't mean others do. Get in tune dude. The pack is UNOFFICIAL AND FREE. WTF is you crying about. You DON'T have to use it. Us geeks are not living by the Microsoft standards. Now spin the @#$% off, before I run you over. The latest browsers wont run under 98 without KernelEx because they are DESIGNED FOR WINDOWS NT. The fact they are DESIGNED FOR WINDOWS NT is the reason we NEED KernelEx. We NEED KernelEx because NT is NOT 100% COMPATIBLE with Windows 9x. .. HENCE ... FILES FROM WINDOWS NT ARE NOT 100% COMPATIBLE WITH WINDOWS 9X! How hard is this concept to grasp? You either can't handle this concept, or purposely refuse to understand it. All you can do is take unwarranted personal offense and behave as if no one knows anything but you...

Look, I really do not want to argue. You and I do not see eye to eye, and I can accept that. But stop acting as if I am attacking you or "knocking" your project. I am simply pointing out a POTENTIAL issue. It IS about a file being new. The fact that it is "new" means that it comes from a NT version of Windows, which is NOT 100% compatible with Windows 9x. The fact that a file is from Microsoft is irrelevant, becuase Microsoft is only interested in supporting Windows NT. If the file can be used under 9x to gain new functionality, that's great, and no one is complaining about that. The problem is that this "new" file WAS NOT DESIGNED for Windows 9x. It MAY work properly uner ALL conditions, it MAY work properly under SOME conditions, and it MAY NOT WORK AT ALL. There is no way to know whether a "new" file works under SOME or ALL conditions without extensive testing beyond the capability of any one individual. It is POSSIBLE that the files in the SP can cause issues. I did not say they ARE, I said they MAY. Hardware is irrelevant in this particular situation. I have no doubt that KernelEx also has issues. I know it does, because I had problems when I tried to use it before as well. This issue MAY be only related to KernelEx, but it MAY be exacerbated by the newer files, and it MAY not. The point is that the newer files introduce another unknown "variable" to the equation.

+1 I really detest Opera... It would be nice if it were possible to pick up the last sources from Firefox 2.x or 3.x and keep it modified/updated it for 9x, but that's far beyond my ability... Yep...

I need to point out something here. I'm not trying to start or restart an argument, and not trying to offend anyone, so please don't take this wrong. Just something to think about. Myself and others have pointed out before that there is a possibility that the various added files in the SP that come from later versions of Windows may not work well under 98SE under all conditions. There may be various hidden bugs and/or other dependencies or "inter"dependencies on components of those later systems that are not present or behave differently under 98SE. Mixing these later versions of files into 98SE, while at the same time running an "overlay" that is supposed to emulate those later versions of Windows may simply be causing these hidden issues to surface, when they might not surface otherwise. The 9x and NT codebases are significantly different. Files not written specifically for 98SE may exhibit strange behavior under certain conditions. I'm NOT trying to say "DONT' USE THOSE FILES," we should be thankful for any later file that can be used in the right situation under 9x. However, one must keep in mind that they were not designed or tested for use under 9x.

See this page on creating INF files... INF files control the packages created by IExpress. Links for IEAK (IExpress) included on the page.

My response was not necessarily directed at you only, it was meant for dencorso as well. It seemed, and the OP obviously took it this way as well, that your responses were to the effect of "you've been given an answer, don't bother trying something else or asking other quasi-related questions." I know you both well enough to know that is not what you meant, but it could be inferred based on the tone of the posts. I agree that the "should still work on 98SE" statement is questionable, but there's no harm in trying. I would also like to see a better, more comprehensive solution to this issue (not to mention the added bonus that a better "generic" driver could be used in Normal Mode on newer systems with no official drivers.) If the file in question was ever provided as part of a redistributable or downloadable package, I must disagree with your statement that we cannot provide it (or at least link to the original package). Back on topic - has anyone tried using the VBE9X driver in Safe Mode? It has some limitations, the worst of which is the inability to use DOS boxes unless they are run in Full Screen mode, but may be a better "overall" solution.

I see no reason why Mikl should not be able to continue his questioning here, or his experiment... From what I read in the other linked thread, the "solution" given is far from perfect or universal, so I don't understand "resistance" toward trying something else...?

Welcome to MSFN! You're definitely in the right place for all of your Windows 9x pursuits.