250 GB Drive with Win98se - problem...

[NewMan]

I recently upgraded my drive to a 250GB Seagate. I have a multi-boot
config with WinXP and Win98se.

WinXP is fine, but Win98se cannot see anything above 137GB.

After some research, I discovered that my motherboard has a VIA
chipset, and it looks like the drivers for Win98se do not support the
48 bit LBA adressing.

I have a Shuttle AV49P/N motherboard. BIOS fully supports the drive,
and WinXP has no problem accessing the full range of the drive.

Does anyone know if there are updated VIA drivers avaiable somewhere
for Win98se to allow full access to the newer large drives???

Thanks

 

[Arno Wagner]

I recently upgraded my drive to a 250GB Seagate. I have a multi-boot
config with WinXP and Win98se.

WinXP is fine, but Win98se cannot see anything above 137GB.

After some research, I discovered that my motherboard has a VIA
chipset, and it looks like the drivers for Win98se do not support the
48 bit LBA adressing.

I have a Shuttle AV49P/N motherboard. BIOS fully supports the drive,
and WinXP has no problem accessing the full range of the drive.

Does anyone know if there are updated VIA drivers avaiable somewhere
for Win98se to allow full access to the newer large drives???

No idea, but why not keep w98se confined to the first 137GB?

Arno

 

[Aidan Karley]

Does anyone know if there are updated VIA drivers avaiable somewhere
for Win98se to allow full access to the newer large drives???

Pretty unlikely since Win98 is now out of the support cycle (and
has been for a year or several now??).
Some drives have in the past shipped with a jumper setting that
can make a single drive device appear on the cable as two devices of
half the capacity. This was, I understand, specifically for this sort
of "1-bit short" addressing configuration. I never played with it on
the first 1GB drive I got (also the first IDE drive, IIRC) ; I just
kept DOS5/Windows 3.1 in the first 540 MB of the drive and put my first
Linuxen in the second half of the disc. Never bothered with it for any
other times when the 524MB boundary was an issue.
I never noticed the 8GB boundary - I had recently got my first
laptop, and the biggest hard drive I could afford for it was only 6GB
(still in use).
There was apparently a 48GB boundary on some laptop chipsets,
but again at that time the biggest laptop drive I could afford at the
time was a sedate 40GB, which is still my biggest laptop drive.
137GB boundary? Why are hardware engineers (and software, for
that matter) not getting several bits (4?) ahead of the curve, and
staying there? It's getting tedious seeing these problems recurring.

 

[Arno Wagner]

Pretty unlikely since Win98 is now out of the support cycle (and
has been for a year or several now??).
Some drives have in the past shipped with a jumper setting that
can make a single drive device appear on the cable as two devices of
half the capacity.

Actually these jumpers limit the accessable capacity. Appearing as two
drives is not possible due to limitations in the IDE bis design.
The last of these I saw were limiters to 32GB for 40GB drives.

This was, I understand, specifically for this sort
of "1-bit short" addressing configuration. I never played with it on
the first 1GB drive I got (also the first IDE drive, IIRC) ; I just
kept DOS5/Windows 3.1 in the first 540 MB of the drive and put my first
Linuxen in the second half of the disc. Never bothered with it for any
other times when the 524MB boundary was an issue.
I never noticed the 8GB boundary - I had recently got my first
laptop, and the biggest hard drive I could afford for it was only 6GB
(still in use).
There was apparently a 48GB boundary on some laptop chipsets,
but again at that time the biggest laptop drive I could afford at the
time was a sedate 40GB, which is still my biggest laptop drive.
137GB boundary? Why are hardware engineers (and software, for
that matter) not getting several bits (4?) ahead of the curve, and
staying there? It's getting tedious seeing these problems recurring.

They are. It is Microsoft that has trouble keeping up. LBA48 was
specified way back. On Linux, e.g., these problems were addressed in
a timely fashion, i.e. before you could buy the too large disks.

Arno

 

[Impmon]

137GB boundary? Why are hardware engineers (and software, for
that matter) not getting several bits (4?) ahead of the curve, and
staying there? It's getting tedious seeing these problems recurring.

Because they never think that far ahead. Waaaay back in the old day,
540 MB was huge and most did just fine with as little as 20MB hard
drive if they ever needed one. Most did OK with itty bitty 320k
floppy disks or a slow cassette that may take up to 10 minutes for a
full program.

Then came complex game with powerful sound and video card. Or some
high end studio needs to do some video working, or artists trying to
render a large complex 3D picture. Add to that bloated Winblows
needing more spaces for the eye candy.

Soon 540MB seemed too small and they implemented a new method that
allowed OS or hardware to handle drives up to 2GB, then to 8GB, 16GB,
and finally 32GB.

Someone was wise enough to bump the next limit from 32MB to 137GB as
that lasted a little while. But nowday it's no longer enough when you
have DV camcorder, DVD burner, and want to handle uncompressed video
in the process to maintain quality.

I don't know what is the current max limit for 48 bit LBA

 

[NewMan]

For most of what I would do with Win98se that is exactly my plan. The
problem is that this is NOT just about the O/S itself! It is about the
data partitions as well!

I have ALWAYS been one of the people who had Win98se installed on
"C:\", and ALL my data on "D:\". I also have ghosted hidden back-ups
of the O/S partitions as well!

As it stands, the "workaround" is that I have a Logical FAT32 data
partition which is below the 137GB limit. This partition can be seen
by all of the O/S in the multi-boot configuration. Then I have a large
data partition above 137Gb that can only be seen by WinXP.

So, operationally, if I want Win98se to see some data, I have to copy
it over to the FAT32 partition before I reboot. Not a big deal, just a
bit of a pain.

And I keep Win98se around for a few things. Firstly I have some old
Pinball simulations that don't run that well on WinXP. WinXP
interferes with the keyboard and causes delays when playing the games.
This usually translates in to the flippers not actuating when you
press the button! So you lose the ball! Very damn annoying.

And I also do some audio processing. The program I have does work on
WinXP, but it also works on Win98se! And the overhead is so high in
WInXP that it is a little annoying. I can do the same operation some
10 times faster in Win98se! On small files, no problem. On a large
project....

Anyways, what I was hoping was that someone might have made a
home-grown driver to solve this problem. If not, I can work around it.

 

[NewMan]

Pretty unlikely since Win98 is now out of the support cycle (and
has been for a year or several now??).
Some drives have in the past shipped with a jumper setting that
can make a single drive device appear on the cable as two devices of
half the capacity. This was, I understand, specifically for this sort
of "1-bit short" addressing configuration. I never played with it on
the first 1GB drive I got (also the first IDE drive, IIRC) ; I just
kept DOS5/Windows 3.1 in the first 540 MB of the drive and put my first
Linuxen in the second half of the disc. Never bothered with it for any
other times when the 524MB boundary was an issue.
I never noticed the 8GB boundary - I had recently got my first
laptop, and the biggest hard drive I could afford for it was only 6GB
(still in use).
There was apparently a 48GB boundary on some laptop chipsets,
but again at that time the biggest laptop drive I could afford at the
time was a sedate 40GB, which is still my biggest laptop drive.
137GB boundary? Why are hardware engineers (and software, for
that matter) not getting several bits (4?) ahead of the curve, and
staying there? It's getting tedious seeing these problems recurring.

Indeed! It looks to me as if the Motherboard manufacturers, for the
most part, have clued in and found a way to let the BIOS properly
identify the actual capacity and configuration of a large hard drive.
Admittedly this sometimes requires a BIOS flash update, but at least
you can do it!

But to have a component of the O/S not be able to recognize what the
BIOS is telling it???? This is a recurring coding snafu! Why in the
hell would anyone with any brains or experience hard code an upper
limit??? This is something that should be dyanmically computed on
startup by the dirver. SO WHAT if it takes a few more seconds to
boot-up???

And I agree with you, you would think that after the problem recurring
more than once in the past that the people coding this stuff would
grab a brain and SOLVE THE PROBLEM. Then again, for Win98se I guess
these people were probably part of Micro$haft, so I guess learning
from your mistakes is too much to expect.

 

[Rod Speed]

For most of what I would do with Win98se that is
exactly my plan. The problem is that this is NOT just
about the O/S itself! It is about the data partitions as well!

I have ALWAYS been one of the people who had Win98se
installed on "C:\", and ALL my data on "D:\". I also have
ghosted hidden back-ups of the O/S partitions as well!

As it stands, the "workaround" is that I have a Logical FAT32 data
partition which is below the 137GB limit. This partition can be seen
by all of the O/S in the multi-boot configuration. Then I have a large
data partition above 137Gb that can only be seen by WinXP.

So, operationally, if I want Win98se to see some
data, I have to copy it over to the FAT32 partition
before I reboot. Not a big deal, just a bit of a pain.

Do you really need to be able to work on all the data when booting SE ?

Likely you dont and you can just keep the data
that is ever used with SE below the 137G limit.

And I keep Win98se around for a few things. Firstly I have some
old Pinball simulations that don't run that well on WinXP. WinXP
interferes with the keyboard and causes delays when playing the
games. This usually translates in to the flippers not actuating when
you press the button! So you lose the ball! Very damn annoying.

So that stuff can obviously stay below the 137G limit.

And I also do some audio processing. The program I have does work
on WinXP, but it also works on Win98se! And the overhead is so high
in WInXP that it is a little annoying. I can do the same operation some
10 times faster in Win98se! On small files, no problem. On a large project....

Sounds like that app is very badly written
or you have stuffed up the config with XP.

 

[Folkert Rienstra]

Because they never think that far ahead. Waaaay back in the old day,
540 MB was huge and most did just fine with as little as 20MB hard
drive if they ever needed one. Most did OK with itty bitty 320k
floppy disks or a slow cassette that may take up to 10 minutes for a
full program.

Then came complex game with powerful sound and video card. Or some
high end studio needs to do some video working, or artists trying to
render a large complex 3D picture. Add to that bloated Winblows
needing more spaces for the eye candy.

Soon 540MB seemed too small and they implemented a new method that
allowed OS or hardware to handle drives up to 2GB, then to 8GB, 16GB,
and finally 32GB.

Nonsense, ...

Someone was wise enough to bump the next limit from 32MB to 137GB

.... it went from 537MB (20-bits) straight to 137GB (28-bits).

as that lasted a little while.
But nowday it's no longer enough when you have DV camcorder, DVD burner,
and want to handle uncompressed video in the process to maintain quality.

I don't know what is the current max limit for 48 bit LBA

Yeah, obviously it cant be 48-bit worth of sectors, right?
That would be too easy.

 

[Folkert Rienstra]

I recently upgraded my drive to a 250GB Seagate. I have a multi-boot
config with WinXP and Win98se.

WinXP is fine, but Win98se cannot see anything above 137GB.

After some research, I discovered that my motherboard has a VIA
chipset,

and it looks like the drivers for Win98se do not support the
48 bit LBA adressing.

And what drivers are these?

I have a Shuttle AV49P/N motherboard. BIOS fully supports the drive,
and WinXP has no problem accessing the full range of the drive.

Does anyone know if there are updated VIA drivers avaiable somewhere
for Win98se to allow full access to the newer large drives???

They are not here: ?
http://www.viaarena.com/default.aspx?PageID=420&OSID=6&CatID=1110

 

[NewMan]

And what drivers are these?



They are not here: ?
http://www.viaarena.com/default.aspx?PageID=420&OSID=6&CatID=1110

Nope!

That driver is for the newer chip sets. I have the 8237, and I am NOT
using the SATA RAID portion of the controller. If you read the fine
print, it says that for "legacy" stuff, the old mini-port IDE V3.20
driver is included.

Just for laughs I downloaded the 5.30a anyways. When I tried to run
setup, it informed me (correctly) that I had the RAID controller
turned off in the BIOS.

Now, the driver for the SATA RAID controller HAS been updated, so my
bet would be that if I turned on the RAID controller, and hooked up a
couple of SATA drives, that the RAID controller WOULD recognize and
access the larger drives. And that makes sense to me, since the
redundency of a RAID array would dictate that you would definitely
want to be able to access larger disks.

Oh well, no biggie. There is no real issue with moving some data from
the upper portion of the disk to the lower portion. I can always
install a second drive if I have to.

 

[Aidan Karley]

Because they never think that far ahead. Waaaay back in the old day,
540 MB was huge and most did just fine with as little as 20MB hard
drive if they ever needed one.

I hinted earlier that I was there. I started with coding forms we
had to post to the DP centre, and got back reels of paper tape to play
on our teleprinter. My first PC had a 60MB hard drive, to which I added
a 40MB HD (controllered-up to 60MB) 5 years later. When I moved up to a
486 I also got a VL-bus IDE controller and then a 1GB drive (I'd had a
good year). But the 386 continued as my main machine. In short - I've
got that tee-shirt.

Then came complex game with powerful sound and video card.

For some people. I got higher powered machinery for computation
and processing geological data. Which is why I continued using the 386
for things like communications, word processing etc while the big
calculations I put onto the 486 and serial-linked the data between the
machines. Couldn't be bothered with the complications of a network.

Add to that bloated Winblows
needing more spaces for the eye candy.

Where's my copy of DOS6.22/Win3.11 install files ...

20/05/2003 00:10 3,417,279 DOS.zip
08/08/1998 14:42 10,804,306 WININST.ZIP

10 megs for Windows, 3 megs for DOS ; stick that on a Zip disc and carry
a boot floppy and any machine at work you can get back to working. Well,
it worked for me for best part of a decade (with other specialist
programs on the Zip disc). No idea what proportion of current Windows is
eye candy. I was playing last month with FreeDOS, which throws a lot of
stuff into a text mode.

I don't know what is the current max limit for 48 bit LBA

48 bits = 4x10bits + 8 bits
= 1024^4 * 256 (decimal) = 1048576^2 * 256 = 1099511627776 *256
= 281474976710656 addresses on the hard drive, each of which will
represent a 512 byte sector of the drive
= 144115188075855872 bytes per device.
= 144,115,188,075,855,872 or 144 petabytes

There are systems that are starting to push the limits of this
addressing space already, but in highly structured contexts.
Specifically, I used to associate with a physicist at FermiLab and CERN
who was working several years ago on systems that would pipeline
multiple petabytes of data per day.

 

[Aidan Karley]

But to have a component of the O/S not be able to recognize what the
BIOS is telling it???? This is a recurring coding snafu! Why in the
hell would anyone with any brains or experience hard code an upper
limit???

To put the large majority of your customers in the situation of
*needing* to buy a new version of your OS in a mere 2 or 3 Moore's Law
cycles (3 to 4.5 years). From that viewpoint, it's not just a technical
choice to program this way, but an economic necessity.

 

[Arno Wagner]

To put the large majority of your customers in the situation of
*needing* to buy a new version of your OS in a mere 2 or 3 Moore's Law
cycles (3 to 4.5 years). From that viewpoint, it's not just a technical
choice to program this way, but an economic necessity.

MS has a history of doing that. A number of DOS updates (well,
new versions only available at full retail price) did nothing
but increase the possible disk size.

Arno

 

[NewMan]

Point taken. And this makes sense considering a recent statement by
Micro$haft that no user would have to wait more than 2 years for the
release of the next O/S!

So, if the next generation in bloatware also requires the next
generation of hardware, then you have a tidy little revenue generator!
And considering that Micro$haft owns part of Intel....

Can you say anti-trust???

Linux is starting to look better all the time!

 

[Splork]

I recently upgraded my drive to a 250GB Seagate. I have a multi-boot
config with WinXP and Win98se.

WinXP is fine, but Win98se cannot see anything above 137GB.

After some research, I discovered that my motherboard has a VIA
chipset, and it looks like the drivers for Win98se do not support the
48 bit LBA adressing.

I have a Shuttle AV49P/N motherboard. BIOS fully supports the drive,
and WinXP has no problem accessing the full range of the drive.

Does anyone know if there are updated VIA drivers avaiable somewhere
for Win98se to allow full access to the newer large drives???

Thanks

The problem is 137GB Disk Total. Partitioning will not help.

There is a 3rd party esdi_506.pdr available but I forget where. I found it
searching for "data corruption on large hard drives" or somesuch.

Package may be named
"Big HDD (>137Gb) support for Windows 98se"

You will have to read the discussions to see if the Via chipset poses additional
problems and to choose the correct version of the driver.

Using W98SE and STORING over 137GB on the drive will cause data corruption of
files and folders elsewhere on the drive.

 

[Folkert Rienstra]

The problem is 137GB Disk Total.

Partitioning will not help.

And where did he say anything about partitoning?

There is a 3rd party esdi_506.pdr available but I forget where.

http://members.aol.com/rloew1/Programs/Patch137.htm

referenced from here http://www.48bitlba.com/tools.htm

I found it searching for "data corruption on large hard drives" or somesuch.

Package may be named "Big HDD (>137Gb) support for Windows 98se"

You will have to read the discussions to see if the Via chipset poses additional
problems and to choose the correct version of the driver.

Using W98SE and STORING over 137GB on the drive will cause data corruption
of files and folders

Actually that depends on whether the original driver is flawed.
If not it will just issue an error message to the calling app.

elsewhere on the drive.

It goes to diskaddress modulo 128GiB.
In other words the first sector over 128GiB goes to your MBR (bootsector).

 

[Rod Speed]

The problem is 137GB Disk Total. Partitioning will not help.

Its more complicated than that. If you partition the drive so that
everything above the 137G limit is in an NTFS partition that SE
cannot see, it wont be able to write there, so normal ops cant
cause any problems when done in SE. So partitioning does help.

There is a 3rd party esdi_506.pdr available but I forget where.
I found it searching for "data corruption on large hard drives" or
somesuch.

Package may be named
"Big HDD (>137Gb) support for Windows 98se"

You will have to read the discussions to see if the Via chipset poses
additional problems and to choose the correct version of the driver.

Using W98SE and STORING over 137GB on the drive will cause
data corruption of files and folders elsewhere on the drive.

Yes, but it isnt hard to ensure that SE cant write above the 137G
limit by having that area formatted to NTFS which SE cant write to.

 

[NewMan]

Unless, of course, you install this:

http://www.sysinternals.com/Utilities/NtfsWindows98.html

I will be trying the other patched file later & will report back!

 

[NewMan]

THANK YOU! THANK YOU!!! THANK YOU!!!!!!!!!!

Your "tip" pointed me in the right direction!

I have located a patched driver file for Win98se that DOES allow 48
Bit LBA!

I downloaded it and installed it in my Win98se O/S, and now Win98se
has NO PROBLEM seeing ALL of the partitions on my 250GB Seagate!

Here is the link:

http://www.msfn.org/board/index.php?showtopic=78592

I downloaded the file: 4102226F.ZIP

This contains a patched version of the file:

esdi_506.pdr

copy the file to the Win98se partition, and then go to DOS mode.

Go to the directory

c:\windows\iosubsys

rename esdi_506.pdr to esdi_506.sav (just in case!)

and then copy the NEW esdi_506.pdr to that directory.

Then reboot! That is all there was to it!

Works PERFECTLY.