LoneCrusader

Yes, I was learnt that USB 2.0 hub contains also speed translator so it can connect USB 1.1 devices. The main difference is that now all ports (and connected devices) goes through the hub and not directly to EHCI. BTW DOSUSB driver should handle this, it also detected mass storage device but failed later on acessing the drive letter. OK, this gives me a hope. I'll try to clean old nUSB and reinstall again. I'm just confused that there's no "!" mark in devmgr on any USB controller but it silently ignore plugged devices. I see only a short flash from LED on USB stick when plugged in. No new device detected, neither unknown device in devmgr... Just a guess... If a motherboard has only USB2 controllers then no USB1 driver files would get loaded on that system. Since Windows 9x and Windows 2K handle USB1 and USB2 devices with different driver stacks it may be necessary for USBD.SYS to be loaded with the USB2 driver (or at least copied to the system) in order to provide complete functionality. If I remember correctly the USB2 stack still uses some calls to USBD.SYS, which could be absent if no USB1 controllers were detected. I believe I saw blackwingcat say something to this effect over in the 2K forum but I don't remember where offhand. PS - I'm getting ready to post the newer Intel INFs over in my other thread.

It's already "finished" for the most part, I just haven't had the chance to write a README on how to use it. An earlier version is here that goes up to ICH9. I simply added all of the Intel device data to the existing MACHINE.INF and MACHINE2.INF, therefore these updated files can be placed in the 98SE SETUP folder and they will automatically be loaded instead of the older version in the CABs. Using this method from my first version works, but soon MACHINE2.INF grows too large and causes errors during SETUP. I had to find a workaround for this so basically I expanded the "MACHINE.INF, MACHINE2.INF" structure based on the multiple MONITORx.INFs. A new MACHINE3, MACHINE4, and MACHINE5.INF are added. One can place these in the 98SE SETUP folder as well as the previous version, however 98SE does not know to automatically copy them to the /INF folder during SETUP. This means they will need to be either; a)manually copied to the /INF folder after the first phase and before the second phase of SETUP, or b)other SETUP INF files such as COPY, PRECOPY, SUBASE, etc will need to be edited to force SETUP to copy them like MACHINE.INF and MACHINE2.INF. I am running a fully slipstreamed build of both 95 OSR2 and 98SE these days so I have not tested them with a "vanilla" installation. All of the "chipset" INFs I consider finished. I have also expanded the USB, USB2, and MSHDC INFs to include newer devices, but these are not so simple to make "generic," for example the USB.INF would depend on whether the target system is using NUSB 3.3 or earlier or 3.5 and later and the MSHDC.INF would depend on whether or not the target system will use rloew's SATA patch or whether all SATA devices will be set to Legacy Mode, etc. My "project building" computer has been down for a couple of weeks, I had to have the motherboard recapped (capacitor plague struck), but I have it back now and I should have it back up and running soon. I will try to work up a README and get it released or if I don't get to it in a few days I will just send you the files. As rloew mentioned I was referring to the patches for 512MB cards. But also we found situations where spurious Option ROM's were causing resource conflicts and causing video cards not to work properly. Sometimes these can be handled with patches as well.

I've been working some with newer Intel boards but I haven't made it that far yet. I've been stepping up bit by bit, looking for issues and solutions. The last board I tried with was a P45/ICH10 chipset board. I have 98SE up and running on it just fine, but for some reason 95 OSR2 doesn't load on this system with the same video card (7950GT) that it worked with on the previous X48/ICH9 system. I am using rloew's RAM and SATA patches and would not even make them attempt without them though... SETUP /P I as mentioned is very useful in eliminating ACPI-related issues and random conflicts in the Device Manager, especially ones on "Motherboard resources." rloew has also recently developed some patches for the 7xxx NVidia graphics cards and we also identified some situations where other patches are needed to handle system-specific conflicts. I have an Intel P67 board but haven't experimented with it yet. Other experiments are on hold until I can devote some more money to the project. I did also create a set of INF's for 9x that is equivalent to version 9.4.0.1017 of the Chipset Identicfication Utility (the last for XP I believe??). They differ from the older ones you created in that they are designed for "slipstreaming" rather than installing with an installer. I have not released them publicly because I haven't gotten around to writing a README to explain their use in a non-slipstreamed setup.

No.. Firefox does have that option. "Save Link As" = "Save Target As"

Read my post #2 in this thread, same applies to your question. Also of note, as far as I know ASUS uses AMI BIOS, which can be very unfriendly to Windows 9x depending on the application. I know Intel uses a modified version of AMI BIOS, and it causes a lot of issues for 9x on Intel-branded boards where 9x was not officially supported. These issues vanish completely on virtually identical hardware (3rd-party boards with Intel chipsets) when an AWARD BIOS is used.

Appears to use AWARD BIOS. That's good for starters. My Slipstreamable INFs will handle the chipset. I recommend rloew's RAM patch to allow you to improve performance by maxing out the RAM on this board. I advise purchasing and using rloew's SATA patch to have fully functional SATA drive support. As MrMateczko pointed out there may be hope for the Network card. No drivers for the onboard video. Does have PCI-E slot for a 98-compatible card. No drivers for the onboard HD Audio. NO PCI SLOT to add a 98-compatible card. I have no experience with USB Audio devices. Intel ICH7/8/9 Chipsets can be handled with my Slipstreamable INF's. The last video cards known to work with Windows 98SE are the NVidia 7950GT and 7950GX2. If the 7950GT is a 512MB version it may require rloew's NVSIZE patches to work properly. The 7950GX2 is a 1GB card and would definitely require the patch but it is currently unknown if the card can be used to its full potential under 98 (as it is essentially a "dual GPU" card). I used to have a couple of SB Live! X-Gamer cards in my older systems. I believe they did implement EAX.

PCI Express works with 98SE in most cases. You will need to use a video card that has 98 compatible drivers though, such as an NVidia 6xxx/7xxx Series or ATI X7xx/X8xx Series card. I haven't looked into ITX boards so I really have no knowledge of them specifically, but I have recently experimented with various post-98 ATX boards. I would recommend that you find one that uses an Intel ICH7/8/9 Chipset BUT is NOT Intel branded! (third party manufacturer such as Gigabyte, MSI, DFI, etc) Be sure that it uses an AWARD BIOS NOT an AMI BIOS as well. Once these conditions are met then you will want to hopefully find a board that uses onboard devices that are friendly to 9x as well, such as the Network card chips and so forth. No HD Audio drivers are available for 98, so you will need a PCI slot to add a 98 compatible sound card. Keep in mind that if the onboard devices have no 98 drivers then you will need more PCI slots to add compatible cards.

Does the power light on the floppy stay on continuously? If so this usually indicates that the FDD cable is plugged in upside down...

One can also simply place the BIOS update files on the hard drive and boot 9x to DOS to run them by pressing F8 at startup and choosing "Command Prompt only" - this is much safer than using a floppy or USB drive IMO...

I don't have any experience with that specific NForce chipset, but I do have two boards with NForce3 chipsets and one board with an NForce4 chipset that I have run some limited experiments on. All 3 have at least one "major" bug with 9x OS'es. The NForce3 and NForce4 boards will not load a driver for a USB Storage device under 98SE if there is more than 2GB of RAM present and a USB2 driver is loaded. This bug does not affect 95, most likely because no USB2 driver exists (without a major backport of 98SE files). ME unknown, but probably the same as 98SE. XP and presumably 2K are unaffected. The NForce4 board has problems with 9x and the old "Gate A20 line" causing very long hangups during every 9x boot (to the tune of you walk away and come back 30 minutes later to see if it's booted yet). This can be fixed with another patch from our good friend rloew. But, once you get past that, I also saw some anomalies using a 7200GS and the 82.69 graphics drivers under 95 with this board. 98SE may not suffer this. My advice would be to avoid NForce for 9x system building, but YMMV.

I stand corrected. Apparently an ICH4CORE.INF for 95 does not exist in the Intel INF packages. Other INFs are present that get used under 95 when installed on an ICH4 system, but the one containing the device you need is apparently 98 and up only. Your system will run fine without any specific driver for this device, but I know it can be annoying to have Unknown devices show up. For a "quick and dirty" solution: -extract ICH3CORE.INF for 95 from the last 95 compatible package -extract ICH4CORE.INF for 98SE from a 98SE compatible package -copy the contents of the [iNTEL_SYS] section near the top of the file from the ICH4 file to the same place in the ICH3 file, example: %PCI\VEN_8086&DEV_24C3.DeviceDesc%=NO_DRV,PCI\VEN_8086&DEV_24C3 -then move to the bottom of the file (bottom of [strings] section) and copy the ICH4 device identifier strings under INTEL_MFG to the same location in the ICH3 file, example: PCI\VEN_8086&DEV_24C3.DeviceDesc="Intel® 82801DB/DBM SMBus Controller - 24C3 " -save the modified ICH3CORE.INF and point your Device Manager to it to update the 24C3 driver. I don't believe any 95 drivers exist for Intel's Gigabit Network cards. I checked the files I have for 95 and no mention of that Device ID.

Export a registry file of HKEY_LOCAL_MACHINE\ENUM\PCI on your Windows 95 Setup and attach it here so I can see what device ID's are in your laptop and compare them to my INFs.

Hmm.. I just looked at the version displayed on the top of the Properties tab. Apparently their "version control" is lacking. lol

Found version 2.11.15.0 here. Just passing through, I'm not familiar with this issue, but when I saw it mentioned I thought it best to try and find any HP tools associated with 9x since they wiped them all at Microsoft's bidding. EDIT: typo

when you say other manufacturers may provide the drivers, is that mean integrated where the operating system loads the sata driver without needing ide legacy / compatibility mode or is it something that you would have to install later for sata devices to be properly configured or some other way that isn't automatically done with the motherboard? also with the sata add on cards that have 9x drivers, does that also have to be configured some different way or do the drivers automatically appear and load with the os install?If a SATA chip/chipset manufacturer provides a driver for 9x, then yes, it can be used to install 9x without the need for Legacy IDE PATA mode but you must know how to integrate the driver into the SETUP process. I don't know if it is possible to set up 9x without either legacy mode or pre-integrating the driver (or using rloew's patch as I do), I have never tried. rloew or others may know more about this. No drivers will "automatically appear" and automatically load with the OS install, unless they were already included in the original 9x CABs. Ideally you need the drivers in place beforehand, whether installing from scratch or adding a new card. Another point in favor of rloew's patch, as it patches and uses the already existing 9x IDE driver.

While "possible" I highly doubt that DOSBOX would be a "desirable" method of running Windows 9x. It's good for DOS programs as intended, but probably VirtualPC or VMware either one would be better choices for virtualizing the whole 9x OS. DOSBOX actually runs under Windows 9x for one, while others are 2K/XP and up only. I haven't used it extensively yet but it does appear to be useful for running older DOS games on newer systems where the CPU is too fast and renders the game unplayable. Best of both worlds? A newer, fast system to run 9x and Windows applications, while having DOSBOX installed to run older DOS programs that don't like the faster hardware. Intel does not provide any SATA controller drivers for Windows 9x. The boards you mentioned are perfectly compatible with Windows 9x so long as you set all SATA devices to Legacy/PATA compatibility mode. The Intel chipset driver packages are actually just text INF's that specifically identify common onboard devices. These are technically not even necessary and a system can run without them. No .VXD, .PDR, .SYS, or any other actual "driver" files are provided. Other chipset manufacturers such as VIA may well provide drivers. I am not as familiar with them as I am with Intel-based systems.

Apparently you've never had a system with an Intel ICH5 or later chipset. While they may be available for some third-party SATA controller chips, drivers are NOT available for most chipset-integrated SATA controllers. This is why rloew's patch is the best available solution, because it handles all possibilities without the need to purchase extra hardware. I depend on them every day. Have for the past 10 years. No signs of any trouble, and no plans to change either.

As far as I know none of these ATI cards came in 512MB versions... It is possible that these "last 9x supported" ATI cards and/or the last 9x driver package have some unresolved bugs with them. ATI only claims official support up to the 9800 XT and doesn't mention ANY X-series card as being supported. I am reminded of the issues I had with a DOS game and the X8xx cards. I don't have any X700 or X600 cards to check for the same issue.

Here. I have it in my slipstream build but without going back and reviewing all my old notes I don't remember how I had to extract it.PCCARD.ZIP

I was surprised that no further issues showed up.. but once the secondary profile was deleted it just booted into the correct one without any complaints. Still bizarre, but at least I know how to fix it now.

Thanks for the link! This procedure from that link worked to fix the problem:

Hello everyone.. hoping someone here can help me sort out a strange XP issue that I'm trying to fix on a friend's computer. First of all I have no idea how the system got into the shape that it's in now, so I have nothing to go on as far as what happened to it. This friend of mine always seems to have the strangest computer problems... I've never seen anything like this; but of course I never set up or deal with multiple user accounts on any operating system and always run as Administrator (and if that statement causes you any distress, I'm not interested in hearing about it). So, I set this machine up for my friend a few years back. The system is multi-boot but only XP seems to be having issues. It's XP Professional with SP3. When I set XP up, I entered my friend's name when it asks who will use the computer, so it associated his name to the "Administrator" account. I did not create any other accounts or set a password since he's the only person using it, so when the system boots it goes directly to the Desktop and does not ask for those things. Now, somehow he has managed to get "shut out" of his own account. I won't say "locked out" because it isn't really locked, there's nothing actively preventing access to the account, it just never allows the option to enter. When the machine boots up, it creates a brand new user account without asking or offering to switch users. This account reverts everything back to looking like a fresh installation... the default theme and background are set, there aren't any shortcuts to his programs on the Start Menu or on the Desktop, there aren't any files in the "My Documents" folders, etc etc. If I try to "Log off" and "Switch user," it only shows up his name and claims he is already logged in. An examination of the "Documents and Settings" folder reveals that his original account name folder, "Harold" shows up as it should, but a new user folder, "Harold.HLG" has been created as well. This folder contains the empty "My Documents" folders and such that I am seeing. If I delete this folder and reboot the same process happens again. "Harold.HLG" does not show up under the User Accounts panel, only the original "Harold". If I boot into Safe Mode and choose Administrator, I am still not able to see any more accounts other than the original "Harold." So, apparently the system believes that "Harold" and "Harold.HLG" are the same account, but how does one get back into the correct "user profile" and escape reverting back to the default new account?

You're welcome. I'm glad I was able to help!

Doesn't make much sense NOT to buy and use Rloew's patch for a setup such as this. What's $10 dollars for working SATA compared to the cost of the rest of the system? Same could be said for the RAM patch, but to each his own I guess. This is board specific. I have two ICH8 boards where USB2 works fine under 98SE. The only issue I have seen with it was on an Intel DP43TF ICH10 board. I think xRayer is the one who provided the other INF method.. unless MDGx had one as well? I have a more complete set of INFs that does support ICH10, but it requires slightly different handling than the set I released before. You may remember I had to trim out some irrelevant stuff from the final version that supported ICH7/8/9 because the INF was too large and caused errors during SETUP. ICH10 and later Intel systems increased the already large INF size exponentially, so I had to split the data into more INFs. MACHINE.INF and MACHINE2.INF can be directly replaced as with the previous set, and 98SE will automatically use these, but the MACHINE3.INF, MACHINE4.INF, and MACHINE5.INF required for all of the data may have to be copied manually or you may have to edit another SETUP INF file in order to have them copied automatically. I haven't tested them with "vanilla" 98SE, but I have used them successfully in my "full slipstream" build. Did you make certain that you had the latest BIOS version on your system? I had to update to the very latest BIOS to have support for the Q9550 on my Gigabyte GA-965P-DQ6.

%24% is a "LDID" reference to a specific directory. I'm not sure which offhand. They are not defined in the strings section. Common Windows directories are assigned a number, or "LDID" (Logical Directory ID) that remains standard and can be used during SETUP or when installing programs. %11% is SYSTEM, %22% is SYSTEM\VMM32, %12 is SYSTEM\IOSUBSYS, etc etc. Here's an excerpt from the INF file for a HotFix. It doesn't define %24% though. [DestinationDirs]; 10=Windows, 11=SYSTEM, 12=IOSUBSYS, 13=COMMAND, 14=Control Panel, 15=Printers, 16=Workgroup; 17=INF, 18=Help, 19=Administration, 20=Fonts, 21=Viewers, 22=VMM32, 23=Color, 25=Shared; 26=Winboot, 27=Machine, 28=Host Winboot, 30=Boot drv root, 31=Root of Boot drv Host; 00=Null (new) LDID, 01=Source drv:\path, 02=Temp Setup, 03=Uninstall, 04=Backup