rloew

True. FAT32 is very flexible and has wide limits. But going above 26 million clusters limits its usability because most of the available programs to give it maintenance weren't written to support so many clusters, as discussed elsewhere: You can double these limits, for Windows SCANDISK only, by using 64KiB Clusters, but you cannot use such a partition as a boot drive or with some older 16-Bit Programs including DOS SCANDISK 128KiB Clusters will raise the limit to the full 2TiB using Patched Filesystem Drivers..

Well, NO. With SOME free stuff you may get that, with some other you can get MUCH more than with paid for solution: it's not the marketing model that makes the difference, it is the quality of the programs and the dedication of the programmers that make a difference. I should have inserted "and/or" in the list. #3 does not apply to all programs. The marketing model does determine whethger and how you pay for the product. There are very few people creating quality free products without the other considerations I listed, so #1 still applies. But we do HAVE NTFS drivers with source: NTFS-3G, and the good guys at ReactOS before or later will release a working NTFS driver. And exFAT is not that different from FAT32, for which driver source is available. Sure it won't probably allow INSTALLING Win9x on such systems, but that would be rather pointless anyway. jaclaz If it can't be installed, why write a driver for it?

These companies/developers did not develop these products to be given out freely. They did one or more of the following: 1. Developed the product when Windows 9X was popular, made their money, then gave it away for free when it was no longer worth marketing. 2. Started Developing the product when Windows 9X was popular, abandoned the effort when the market collapsed, then gave away the Betas. 3. Released a crippled version to promote a pay version. 4. Developed a product for a more marketable OS such as XP/Vista/7 and released a free version for 9X for publicity. 5. A few people do it for their 15 minutes of fame or to promote other paid products. Anyone who takes up your challenge would not be covered by reasons 1-4 above. At best only #5 would apply. I write my programs specifically for Windows 9X in the current market so reasons #1, #2, and #4 do not apply. #3 applies to my Demo Versions of some products. Few of my products have tie ins to other products or more popular OSes so #5 doesn't apply. I also provide support and updates for my products. Few freebie publishers will go through the trouble of providing support or updates. As the old saying goes: you get what you pay for. With free stuff you get: 1. A limited selection. 2. Most are low value unmarketable products. 3. Incomplete or buggy programs. 4. Little or no support or updates. 5. Adware, spyware, spam or other scams. 6. Come-ons for paid products or services. 7. Illegal copies. Maybe, or maybe having a "C based" tool, would make it easier. I wrote my VxDs in pure Assembly to make them more streamlined and efficient. There is no reason that a small Assembly module cannot be combined with C Code to make a VxD. The original examples from Microsoft, I used to start writing my VxDs, are mixed C and Assembly Code. Using C for a larger project would be easier, but most of the work in porting an existing Driver is reverse engineering the undocumented binaries.

I recently purchased a 1TB Buffalo DriveStation FlexNet Model HD-CE1.0TLU2 External USB/NAS Drive. The USB works fine. Under Windows XP it works fine. In Windows 98SE I can see the Share using NET VIEW only by using the Server Name not by it's IP Address. NET USE without a Drive Letter will let me connect to it, but fails if I specify a Drive letter. It reports Invalid Network Type Error 66 and tells me to specify a Drive Letter which of course I already had done. My Amiga Computer reports an Unrecognized Protocol. I called Buffalo and was told it does not support Windows 98, Macs or Linux. Apparently another class of Hardware is losing support for Windows 9X.

As often happens, OFF-TOPIC, but not much : How much money do you think these guys are making nowadays from selling this?: http://www.techsoftpl.com/vxd/index.htm at US$ 429 apiece? http://www.techsoftpl.com/vxd/order.htm Maybe if they would decide to "let go" that software or provide a "complimentary copy" a C programmer willing to take the challenge could even be found. jaclaz I doubt a free copy of VxDWriter would get meny people writing VxDs for Windows 9X. I already write VxDs for Windows 9X without needing VxdWriter. I have written a 64-Bit Memory VxD and am in the process of writing a Multi-Core VxD for Windows 9X. Porting an undocumented Driver would be a lot more work. I haven't seen enough interest in my exisiting VxDs to justify the much greater task of porting ExFAT or other XP Drivers.

Hmm... So is Int30 set as a real mode interrupt or as one in 'protected' mode (i.e via a trap/intterupt gate)? In V86 mode, all Interrupts switch to protected mode. The Interrupt handler can return to V86 mode as if a Real Mode Interrupt occurred, if necessary. It does whatever is needed for the intended function. Contexts can change if the function has to wait. The final IRET automatically switches back to 16-Bit V86 mode. Since my RING 0 switch is not an intended INT 30 function, I have to unwind the System Stack before the IRET.

INT 30 is one of two ways I have found that Windows uses to allow V86 code (16-Bit Programs and Real Mode Drivers) to call Protected Mode functions. Of the four methods I use to enter RING 0 from 16-Bit Code in my RAMDISKs, this is the only one that does not require prior setup or kernel-mode files.

If your motherboard was made in 2001, then there is a good chance that the BIOS does not support Hard Drives larger than 137GB. If so, you risk corruption when: You actually place data beyond the 137GB limit. and 1. Run SCANDISK from DOS (as occurs after a bad shutdown in 98/SE). 2. Run Windows in Safe Mode or with Protected Mode Drivers Disabled. 3. Place files used during boot above the 137GB limit. and/or 4. Install/Unistall Software that has to reboot to complete. You will need a DDO Program, such as my BOOTMAN Package, in addition to the Intel Application Accelerator, to safely use that 250GB Hard Drive. This corruption risk also applies to DOS and Windows XP/Vista/7. To determine if your BIOS Supports Large Hard Disks, try the following: Boot your Computer. If the BIOS reports that the Hard Drive is 137GB then it does NOT support Large Hard Drives. If it reports the correct size, it is probably OK. EPIA motherboards are a significant exception to this rule. To be sure, download the Demo of my High Capacity Disk Patch and run the 48BITLBA.EXE Program from true DOS, not a DOS Box.

But... but... but it isn't enough to fix the MBR partition table format (in which the last two entries, "Sectors Preceding Partition a.k.a. LBA of 1st Sector" and "Number of Sectors in Partition" are both 32-bit numbers, and hence limited to 2 TiB)! You'd have to first fix the PBR (= Boot Record) format, because the last entry in the BPB ("Sector Number", for large volumes) is also a 32-bit number. And while you can have a working disk without a MBR (= "super-floppy"), you cannot have a working disk without a PBR. In any case, your first approach seems safer. But both, to be generally usable for those who multiboot, like me, would require also a Win 2k/XP counterpart, to avoid problems, isn't it so? In the second design: I redefined starting sector entries, in both the MBR and VBR/PBR, in the last 8GiB of the 2TiB range as mappings into 64-Bit Sector space. When encountering one of these starting offsets, my modifications to VFAT.VXD remap the starting offset before adding the partition relative sector number to it. There would be problems mixing OSes since the others would see overlapped partiitons starting near the end of the 2TiB range wrappnig around back over the start of the drive. The first design avoids problems with multiboot since any unmodified OS would not be aware of the additional Partitions in the area above 2TiB. The other OSes would not have access of course. They would have full access to Partitions in the first 2TiB. The disadvantages of this approach are as follows. 1. Problems begin to appear when there are more than 8 physical drives. The Pseudo Drives are treated as physical Drives. This would limit Interrupt 13 compatable Disk Drives to a total of 16TiB. The second approach is only limited by the number of Drive Letters, which is 48TiB (52Tib if A: and B: are repurposed). Increasing the size of a Partition above 2TiB is more complicated and is yet unsolved. 2. Drive letter issues may appear, depending how the Pseudo Drives are setup, as each one's first Primary Partition will be scanned before any extended Partitions. Neither design currently supports booting from Partitions above 2TiB.

Could you tell something more about 2 TiB patch? The 2 TiB Patches are designed to allow the use of Hard Drives larger than 2TiB with Windows 9X. Since no such drives are available yet, I had to simulate a drive to test them, so they are both still in Alpha. The first approach remaps a Hard Drive into smaller Pseudo Drives that are individually Partitioned and Formatted. The second approach extends the MBR Partition format and modifies VFAT.VXD to process it.

Watch those typos! 2GiB and 1GiB should be 2TiB and 1TiB respectively. The 2 versions of the 2TiB Patch both modify ESDI_506.PDR so they will not work with USB. The 1 TiB limit is due to a flaw in VFAT.VXD that may or may not appear depending upon the alignment of clusters and the page sized blocks (4KiB) used for I/O. The problem generally appears as a system freeze when trying to read directories located above the 1TiB boundary. Fortunately, corruption does not appear to occur. The Patch corrects the code in VFAT.VXD. No other files are modified. Note: The 1TiB limit applies to the size of a Partition, not the total size of a Drive. I haven't tested my 1TiB Patch with an USB Drive yet, but it should work. It better, I already started using a 2TB USB Drive for Archiving.

There is no 137GB limit for NUSB. There is no 137GB limit for USB. Period. I'm positive about this. USB itself has no limit, but some old Hard Drive enclosures might be limited to 137GB if they do not support 48-Bit LBA. This should not be a problem if you buy an already packaged Hard drive, but could be if you add or replace the drive in an enclosure not rated for the new size. I have an experimental package to break the 2TB limit but it is currently does not support USB.

They streamlined the process significantly since the last time I asked for a Hotfix. I downloaded and extracted the 4 versions of the CDFS.VXD File. My Patch worked on all of them except the Original Windows 95 CDFS.VXD Version 952.

Another approach to sharing the C: Drive is to install Windows 9X into a different directory (C:\Win9x in Glaurung's method above). When the Computer reboots, boot to DOS. Rename the PROGRA~1 file to PROG9X~1. In the Registry and in WIN9X\SETUP.INI replace all occurances of the followings: PROGRA~1 with PROG9X~1 "Program Files" with "Prog9XX Files" (replace whether quoted or not) Continue Installation. Windows 9X should now be installed entirely in directories that don't conflict with Windows XP. This approach breaks a lot of defaults so extensive use of this installation of Windows 9X is not recommended.

You need to use Safe Mode to remove the problem driver. If you boot in Normal Mode with the SATA2 Drive disconnected, the Device Manager will not show the SATA2 Driver, so you will not be able to remove it. My Windows 98/SE/ME RAM Limitation Patch will give you full support of your 2GB of RAM in Normal Mode and in Safe Mode. The Demo Version will let you run for 10 Minutes. Install the Patch before doing Step 1 in my earlier post. If you use the Demo Version, uninstall the Patch before going to Step 6. There is no conpatability problem, specific to SATA2, with Windows 9X. Some Chipsets have a problem with it. My SATA Patch patches the Hard Disk Controller to deal with Interrupt and Initialization issues. An Associated INF File sets the proper configuration. Note: Officially there is no such thing as "SATA2". It is an unofficial designation for 3GB/Sec SATA drives.

Try the following: 1. Boot into Safe Mode. 2. If you cannot, stop here. 3. Go to the Device Manager 4. Under Hard Disk Controllers, delete all copies of "Standard Dual PCI IDE Controllers" 5. If none, stop here. 6. Reboot into normal mode. 7. If you cannot, stop here. 8. Let it install drivers and reboot. 9. If the problem reappears, you will need my SATA Patch.

OK, but I had set IDE legacy mode in setup (drives was on port 170/1F0, IRQ14/15) that you cannot recognize it is not normal IDE. I didn't need native SATA support. There was dual PCI standard controller in device manager but not working. Maybe it was collision of native SATA. I tried to disable it in device manager but didn't helped. When I tried to remove this dual PCI controller it freezed. How many "Standard Dual PCI IDE Controller" entries are there? If more than one, a Native SATA Controller was recognized incorrectly, in addition to the Legacy Mode Controller. You may need to go into Safe Mode to remove the entry. If you added the PCI ID for the JMicron to an .INF File, it will be recognized as the INF directs rather than as an Unknown Device, even if it is incorrectly configured.

I tried to patch INF files. I filled PCI dev IDs according to latest INF files from intel's winxp/vista package. It also detected native SATA controller correctly but ! mark was around standard dual fifo ide. Anyway I have no longer this board available so cannot perform further tetst. But if someone dosn't want a headache I recommend to buy ICH7 equipped mobo. Adding Device ID's to the .INF is not enough. Native mode requires other changes to the .INF file and a Patch to the ESDI_506.PDR File.

Probably. I don't have the hotfix files to test. If the Patch Program does not complain when Patching, it should work.

A Patch is required to properly support SATA Drives as well as a modified .INF File. The JMicron Driver may be reporting as a RAID Driver. I have a separate .INF File that supported a JMicron SATA RAID Card.

Built in USB 2.0 on 98FE? What drivers did this use? Did the 1.1 card use native USB drivers or did it come with its own? The USB 1.1 Card used the native drivers. I installed the built in USB 2.0 using the Motherboard CD, The Drivers which may have been USB 1.1 only and may have been the native drivers as well. I no longer have it setup. I tried the various generic Drivers with the machine with the USB Card. I haven't tried NUSB with either. I agree. Further research is warranted. I have noticed problems with the default partitioning of USB Keys. I usually repartition them with my RFDISK Program so they are bootable.

Maybe it's Windows Explorer under Win9x, not USB, that is causing the problems. The problem of the sluggish file deletes with Win98 Explorer shows that it is seriously flawed.For moving files, I try not to use Windows Explorer under Win98; I am using Beyond Compare instead, or I boot into WinXP to do the file moves there with WinXP Explorer, but this has lead in a few rare instances to filename problems between Win98 and WinXP, for files containing some foreign characters (e.g. mp3s downloaded with eMule). Also, Win9x Explorer sometimes cannot delete some illegal files, I have to go into WinXP to delete them. I used a Program for managing files called XFILE from a DOS Box. I did not use Windows Explorer. Under 98SE I have not had corruption on the Hard Drive, but I have had transfers freeze leading to errors on the USB Drive. Substituting some files with ME versions seemed to help but other issues caused me to restore the originals. Further work is needed to determine if this would be useful. Sounds like what I observed. I was able to repair the damage using my own tools and had backups for most things. The big problem was AOL's filing cabinet which is constantly updated while using AOL. The active part of the file was fragmented all over the destroyed FAT sectors. I had to match up a dozen fragments out of 128 clusters. One machine was an old computer with a USB 1.1 Card. The other has built in USB 2.0. One time it occurred while I was transferring a few dozen digital photographs to an USB Key. As indicated above, I am not the only one who has seen this problem. #3 which includes #1 and #2.

I have used generic USB drivers with 98FE, but I found a very serious problem. When I transfer large amounts of data from my Hard Drive to a Flash Drive, something gets damaged in the system, causing garbage to be written to the Hard Drive and sometimes the Flash Drive. Usually a Read/Write Error Blue Screen appears, but not until a Directory Sector and Both Copies of one FAT Sector have been destroyed. This has led to some rather tedious recovery efforts. I have observed this problem on two entirely different Computers, using at least 5 different USB Drivers. Mixing 98FE and 98SE files didn't help or didn't work. It may be safe to READ from a Flash Drive but I would not guarantee it. I would not recommend using USB with 98FE. Can't stand the stuff.

If you are referring to my Windows 98/SE/ME RAM Limitation Patch, you can tell by using the following procedure: Run the Patch program (Full or Demo). If the Patch is already installed, the Program will offer to uninstall it. There are a few situations, such as using a patched CAB file, where it may report that the cannot install the Patch. If the Patch is not installed, the Program will Install the Patch. Rerun the Program to uninstall the Patch, if desired.

The reason XP, and the Patched Win9X, only see 3.2GB (in your case), is that the BIOS reserves the rest of the 32-Bit Address space for Memory Mapped I/O and the BIOS ROM. In addition, the BIOS will map AGP apertures and other large PCI card I/O Spaces to this area as needed. Rather than wasting the RAM Space, the BIOS remaps this RAM into the 64-Bit Address Space above 4GiB. Windows 9X can use more than 100MB of RAM internally, if it is available. File Caching can use 512MB alone. Even if an Application is limited to 2GB, running two at the same time brings you up to 4GB. Using RAMDISKs for Swap and/or Temporary Files provides additional speedup. On one machine, I have a 1.5GB RAMDISK that holds my Temporary Internet files. I can download 2 CD sized ISO files simultaneously. In addition, all files are cleared upon reboot.