I have the 95 OSR1 OPK CD you refer to as the OEM 95/95a install disks don't have the OPK on them unlike OSR2. I had made the assumption there was no separate OSR2 OPK CD as the OPK files (although not as comprehensive as the OSR1 OPK) are on the OSR2 setup disk, I have not seen any info to suggest an OSR2 OPK disk exists (until your message), and back in the late 90's when I worked for an OEM (a small DSP one, not Royalty OEM) we were instructed to use the MKCWRT.EXE file on the OSR2 disk to prepare pre-install setup. Happy to be corrected on this of course, but have you or anyone else seen pictures of an OSR2 OPK disk?
Greetings! I found a copy of a Hungarian version of Windows 95 OSR 2.1, but unfortunately some dumbass messed with the ISO so it is definitely not untouched. The damage that as been done is 3 unnecessary files added. AUTORUN.INF, a SETUP.ICO which is a Windows 7 logo and has nothing to do with Windows 95 and a KEYGEN.EXE with also must not be there. All of these have a date of 2009. The second mocking with the image is the ULTRAISO 9.2.0.2536 fingerprints that the person left, because for some reason it was made to be bootable which it shouldn't be in the first place. At least we know CDIMAGE 2.28 was used by Microsoft just like the Finnish version. My question is would it be possible to restore it to an original state? Unfortunately I don't have an untouched image or at least a disc to dump it so that we can compare. Tried to use those hex edited CDIMAGE versions that has versions changed to 2.28, but I was unable to generate an ISO file with the same hashes for the Finnish version of OSR 2.1. I got the size right, but has was still different. Here's a link to the modified image: mega.nz/file/hF81nBAb#yQ_Ip-0053_kRIm9GYfeu2lE-PnTgU_lCC9g5Jb-G-8
@P4R4D0X Thank you for the osr2.1 Ofc, but question is: how to verify it? Do we have a verified image dump/checksum to compare against? Because just remake and call it original is kinda guesswork. Image works and boot capabilities is just a bonus. The keygen embedded doesnt harm either, it's not really a keygen, only a list of serials embedded in a gui. Imho it's better to keep it as-is for now.
I wanted to recreate first the Finnish version of OSR 2.1 from here: https://forums.mydigitallife.net/th...osoft-windows-9x-me.54014/page-3#post-1793747 But I wasn't able to recreate it with the same hash. Like I said. That is also made with CDIMAGE 2.28. The size is the same, but for the love of god I cannot get the same hash. Currently this is the only OSR 2.1 Hungarian version we have. Nothing to compare against with. That's why I asked if it is possible to recreate it that are similar to this one. I agree that this should be kept as-is, because there's nothing better available for now.
@P4R4D0X First off all, going through the posting, I have no reason at all to believe the Finnish version of OSR 2.1 is fake. For this image you "should" be able to recreate with Code: cdimage.exe -lFINW95B_USB -t5/1/97,12:14 FINW95B_USB FINW95B_USB.iso -n TZ byte is 0x04, that's UTC+1 winter time There is one file that does not cooperate: WINCHAT.GID I have looked through OSR 2.1 releases and specific to Scandinavian releases: this file seems to be unique to the Finnish image only. I get one wrong bit related to this file in the least significant nibble at 0x24137, expected value 0x01, cdimage gives 0x00 That's all. It's mind boggling, no clue... According to the ISO standard, this nibble refers to a file attribute. (with in total 8 different values used)
Thank you! I found where I did the mistake with the CDIMAGE command. I didn't include the -n parameter at the end. With that I found the exact same thing with the Finnish image. However now I ditched that and I experimented with the Italian OSR 2.1 image: FAF699EE3B6A4BEE2A180B8827BE6F247DF6A255 I was able to recreate that successfully. The only difference was the TZ byte. I corrected it to 0x04 and the file is now identical and I now get the same hash. Command I used: Code: cdimage.exe -lITNW95B_USB -t5/1/97,12:14 ITNW95B_USB ITNW95B_USB.iso -n With this knowledge I recreated a Hungarian OSR 2.1 image which I cannot verify for now. I corrected TZ byte to 0x04 just like for the Italian image and I got the following hash. Code: cdimage.exe -lHUNW95B_USB -t5/1/97,12:14 HUNW95B_USB HUNW95B_USB.iso -n Code: CRC-32: C2133F80 MD5: A788A8AEEEFF67B9108C4FB817DFF10E SHA-1: CDC919C8D0A7699F1FE803FCBCA864004D21640E What do you think? Definitely this needs a dump from a CD to be 100% sure this is correct and the hashes match. Until then take it with a grain of salt. Edit: I was able to recreate the Italian OSR 2.1 image 1:1 on GMT+1 with no daylight saving and got the same hash and then did the Hungarian OSR 2.1 and got the hash above.
I haven't seen pictures of a CD, but there's a partial dump from old warez archives of the OPK for a win98 beta (I think beta 2?), the documentation included comes from OSR2 OPK. Also pisetup included with some win95 osr2 recovery media (i forget which).
@P4R4D0X Cool - my thoughts: Both images (the bootable and the one you made of the Hungarian image) works but without a confirmed dump this is still not an image I would consider genuine just because it's remade with cdimage. It "could" ofc be correct but with no checksum from a confirmed dump it's just to be considered as homebrew. That's my 2 cents. On that subject and to enlighten further, I solved this bugger: Step 1 - run this on WINCHAT.GID Code: attrib.exe +H WINCHAT.GID Step 2 - added -h switch to cdimage syntax Code: cdimage.exe -lFINW95B_USB -t5/1/97,12:14 FINW95B_USB FINW95B_USB.iso -n -h TZ: 0x04 WINCHAT.GID - expected file attribute is value "hidden" and cdimage with -h includes hidden file(s) This is an example on how files can sometimes be "not what it seems to be" - this should be observed in general. There is no other images atm to confirm this WINCHAT.GID that i'm aware of. This attribute could be correct, there is also a possibility that this is not correct through bitflip/bitrot but: I have no reason to believe it's not correct. Important thing is that we have a confirmed dump and checksum, more unique dumps is ofc better. Happy New Year folks.
@P4R4D0X very nice indeed. You served us this one on a silver platter Good pictures, overdumped image, some additional background info. I will add this to Good re-dump, nice research. Thank you kindly for bringing another piece of history back to surface. Kind regards
Hi @P4R4D0X I can not check, I'm "proud" to say i have no access to Beta archive's server... It's not for everyone... If you can/are willing to upload original image we can for sure look closer. (can/willing is subjective, I dunno exact policy of Beta archive and individuals, so, no demands my friend)