BetaArchive is a non-profit site and your donations help to keep us online and thriving. If you can, please consider donating a small amount to help keep us online. Donate and find out more here. Thank you! - Andy & mrpijey
Download requests and offers should be made in this forum. Do not request a download if you have under 10 posts. You will be ignored.
Forum rules
Please read the following rules before posting a download request in this area:
1. Don't post a request if you have under 10 posts as stated in the rules. If you do anyway, it will be deleted without further notice. 2. Requests for anything against our rules will not be entertained and you will be warned. 3. Check that we don't already have the file on our FTP servers by using the database linked in the navigation.
This has made it to the "olddos" Naver Cafe recently, on proper floppy images. Note that its included Korean display driver (HECON.SYS) doesn't work correctly on any emulator or virtualizer, so I've attached (but not uploaded to the FTP as it's outside the FTP's scope) a correct disk 1 using VHCON.SYS instead of HECON.SYS - both the original correct disk 1 as I got it on Naver, as well as my own improved version that actually installs VHCON.SYS instead of HECON.SYS and even configures the installed OS to use it.
Here are several screenshots, from Setup to the installed OS (using my improved version of the corrected disk 1 with VHCON.SYS), taken with 86Box (formerly known as PCem-Unofficial) and its integrated screenshot-taking feature (click on a thumbnail to see the full-sized screenshot):
Edit:
- http://citadel.rol.im/tg50.IMA ; an alternate boot disk with slightly different (but still original) files of Korean MS-DOS 5.00, also TriGem OEM, but earlier (October 1991), incomplete, also from the same Naver Cafe.
What I find interesting is that both copies' MSDOS.SYS default to code page 934 (IBM DOS Korean) with the correct parameters for that code page, yet all the strings are in the Wansung/KS-5601-1987/EUC-KR code, which is the encoding also used by Korean DOS 6.x and later where it is given code page number 949. I also noticed that they are variations in string translations among both Korean 5.00's, as well as the 4.01 of which we only have screenshots (binaries deleted by Naver for apparent malware) and the 3.21 of which we also only have screenshots.
This leads me to hypothesize, that official Korean DOS versions from Microsoft and IBM before 6.x used code page 934, but the OEM's retranslated the strings on their own, using Wansung encoding, but never changed the parameters in MSDOS.SYS. This probably led Microsoft and IBM to both give up on code page 934 and just adopt the Wansung encoding as code page 949 (and Johab/KS-5601-1992 as code page 1361). However, Microsoft's departments outside South Korea were probably never made aware of that change, hence why MS-DOS 6.22's COUNTRY.SYS still mentions 934 as the Korean DOS code page.
However, to solve this puzzle once and for all, we need the retail version of MS-DOS 5.00 Korean, which was also released on the Microsoft Far East SDK, as well as any Korean IBM DOS that's 5.0x or earlier.
Edit: And to make it clear why the puzzle exists - code page 934 is completely different from both 949 and 1361. In fact code page 934 seems to be a derivative of Japanese code page 932 (Shift-JIS) modified for supporting primarily Korean rather than Japanese.
I remember that when I e-mailed IBM some years ago for information about code page 934, they forwarded my request to their Japanese department rather than the Korean one, and that department then sent me the unicode mappings for the code page. So I suspect that the code page was created at IBM Japan as a derivation of Japanese code page 932 but the South Korean government as well as OEM's most likely refused to adopt it for political reasons, instead standardizing the Unix-originating EUC-KR as KS-5601-1987/Wansung code and retranslating MS-DOS 5.00 on their own in that encoding, most likely creating a mess, which probably led Microsoft and IBM to adopt KS-5601-1987/Wansung code as code page 949 and even the later KS-5601-1992/Johab code as code page 1361 (and both Microsoft's and IBM's HBIOS from DOS 6.x onwards support both Wansung and Johab codes, but not code page 934).
Edit #2: And there's a similar situation regarding Traditional Chinese code pages too. IBM DOS 6.x still use code page 938 which is IBM's code page, MS-DOS seems to never have existed in Traditional Chinese (not even the DOS included with Windows 9x was localized into Traditional Chinese), and Windows ended up adopting code page 950 which was from what I know based on E-Ten's encoding. Code page 938 also seems to be derived from code page 932, though it differs from it much more than 934 does.
Interestingly enough, Simplified Chinese code page 936 is not derived from code page 932 at all, and is instead developed according to the GB standard formally adopted by the People's Republic of China, which is in turn based on the Unix-originating EUC-CN. Also, code page 950 seems to at least resemble EUC-TW, but I need to investigate that further.
Battler wrote:This has made it to the "olddos" Naver Cafe recently, on proper floppy images. Note that its included Korean display driver (HECON.SYS) doesn't work correctly on any emulator or virtualizer, so I've attached (but not uploaded to the FTP as it's outside the FTP's scope) a correct disk 1 using VHCON.SYS instead of HECON.SYS
Btw what's the main difference between those two (seems like VHCON.SYS comes with additional font) ?
I'd suggest that it 's MS-DOS 5.0a (dated March 1992, contains EMM386 v4.33)...
Edit:
- http://citadel.rol.im/tg50.IMA ; an alternate boot disk with slightly different (but still original) files of Korean MS-DOS 5.00, also TriGem OEM, but earlier (October 1991), incomplete, also from the same Naver Cafe.
So I started QBASIC in this (EDIT is also affected, because it uses QBASIC as the base), and I saw this:
Notice how the scrollbars are using characters with ASCII codes below 32, exactly as would be expected from a piece of software designed for code page 934 (which has certain box drawing characters there, much like code page 932 does). Evidently, when TriGem were retranslating and reencoding the strings and modifying everything else to suit the Wansung encoding, they forgot to modify the code for the scrollbars in QBASIC.
