VOGONS


Reply 20 of 21, by Riikcakirds

User metadata
Rank Member
Rank
Member
AaronS wrote on 2024-03-28, 16:16:
Okay MSFN is back online, so I downloaded the two ZIP files that are labelled as v1.1 (4102225F & 4102226F) http://www.msfn.org/ […]
Show full quote

Okay MSFN is back online, so I downloaded the two ZIP files that are labelled as v1.1 (4102225F & 4102226F)
http://www.msfn.org/board/topic/78592-enable4 … -137gb-barrier/

I SHA1 checked both and neither of them matched the one I was using from the bhdd31 pack. Checking the dates it actually matches the "4102225F" one but as I said, with a different checksum. There's also "4102226F" but it has this note attached:
"only for replacing an existing v4.10.2226 file"

but anyway if you could let me know which one you're using, it seems like I was maybe using a bad ESDI_506.PDR file.
4102225F - July 20th 2006, 1:38pm - 232cf58f76d44851e1888138e30ab5e501353723
4102226F - July 29th 2006, 3:02am - 125ad17b11363f761e4f0b1e0dacbecebe846aaf

This thread sums up all the differences (its 3 pages)
https://msfn.org/board/topic/141244-question- … from-microsoft/

Pretty confusing but to sum it up - 4.10.2225 and 4.10.2226 are the official versioning numbers by Microsoft of updates released for esdi_506.pdr for Win98se.
Version 4.10.2230 is a made up number by a user on the MSFN forum called Maximus-Decim. It is just LLXX unofficial 48bit modified version of 4.10.2225 with a hex edit to change the version number to 4.10.2230. If you hex compare them they only have the version numbers changed in two places. You may want to check this in detail to verify on the link I posted above.
Also of note is a user called ' Fredledingue' on that link above mentions that they could only get the 48bit modded version of 4.10.2222 to work, so you may want to try that one as well.

I personally was using 4.10.2230 - SHA1 - a10c619e214f87da082cd0b0905e6613c689974f and it worked fine.
I later applied some Winme updates to my 98se install, and can't remember exactly why now but I have been using 4.90.3000 - SHA1 98abad358e29839bb042988bbc03adefc336c999 since then. It also works fine on my 1TB drive, no corruption. You could try that version to.
You can get the Winme 4.90.3000 version from here:
http://web.archive.org/web/20210506185314/htt … mdgx.com/files/
the file is called ME48BLBA.EXE, open it with 7zip.

Just One last thing, when you copied the files in the earlier post booting directly to real mode dos 7.1, did you also check the hash of the original 20 files again, not just the copied ones? Sometime the corruption can affect parts below <137GB when you write above it.

Reply 21 of 21, by AaronS

User metadata
Rank Newbie
Rank
Newbie

It's been a long while since I last looked into this, I since acquired some other laptops and other SATA>IDE adapters, same problem on those too. All these laptops have one thing in common, they all use Intel Chipsets, the Thinkpad T42 uses 852PM + ICH4-M, another uses 845PE + ICH4 and another uses 848P + ICH5 (its too bad this one doesn't have a SATA port instead since ICH5 was the first to support SATA I believe). So narrowing it down to just being an Intel thing, unless you guys have had success on Intel Chipsets? I've also had other issues on the latter two laptops not liking SATA to IDE adapters entirely, though I'm 99% sure this is a BIOS thing with the UDMA2 thing.

Riikcakirds wrote on 2024-03-28, 20:55:
Pretty confusing but to sum it up - 4.10.2225 and 4.10.2226 are the official versioning numbers by Microsoft of updates release […]
Show full quote

Pretty confusing but to sum it up - 4.10.2225 and 4.10.2226 are the official versioning numbers by Microsoft of updates released for esdi_506.pdr for Win98se.
Version 4.10.2230 is a made up number by a user on the MSFN forum called Maximus-Decim. It is just LLXX unofficial 48bit modified version of 4.10.2225 with a hex edit to change the version number to 4.10.2230. If you hex compare them they only have the version numbers changed in two places. You may want to check this in detail to verify on the link I posted above.
Also of note is a user called ' Fredledingue' on that link above mentions that they could only get the 48bit modded version of 4.10.2222 to work, so you may want to try that one as well.

I personally was using 4.10.2230 - SHA1 - a10c619e214f87da082cd0b0905e6613c689974f and it worked fine.
I later applied some Winme updates to my 98se install, and can't remember exactly why now but I have been using 4.90.3000 - SHA1 98abad358e29839bb042988bbc03adefc336c999 since then. It also works fine on my 1TB drive, no corruption. You could try that version to.
You can get the Winme 4.90.3000 version from here:
http://web.archive.org/web/20210506185314/htt … mdgx.com/files/
the file is called ME48BLBA.EXE, open it with 7zip.

I tried all of these and same thing except the Windows ME one, it boots but then locks up rather than blue screen. That said I only replaced the file after several reboots, driver installs etc. rather than right from the first reboot during install, I could try this one again from a fresh install.

Just One last thing, when you copied the files in the earlier post booting directly to real mode dos 7.1, did you also check the hash of the original 20 files again, not just the copied ones? Sometime the corruption can affect parts below <137GB when you write above it.

I didn't. I just did this test again on the 848P+ICH5 though (I used DOS 8.0) and no corruption. Also the latter two laptops even have a "LBA Mode" in the bios which can either be disabled or set to auto, and I imagine when these laptops released, 120GB+ 2.5" drives were starting to hit the market and Windows XP was a few years old now. So it's either a Windows 98 thing or an Intel thing. I guess I could try installing 98 + replace ESDI_506.PDR, and before installing any drivers at all, fill up the drive, I need to install stuff to add/edit the registry thereafter, but don't install the chipset drivers and see what happens. I also read this at MSFN regarding an updated IO.SYS from 2001:

https://msfn.org/board/topic/113142-install-w … #comment-775004

I tried replacing this too but no avail.