This sure is a Very Finnish Project. (Definition: a completely random Finn, always carrying a barely noticeable smirk, is taking an utterly improbable or completely unnecessary project and finishing it to absolute, ridiculous perfection. Warmest greetings to all Finns from the other side of the Gulf! I love you guys forever.)
The first thought that came to my mind were actually the Future Crew, and one of their members were also named Sami (but seems unrelated to this author).
Tl;Dr: "lEEt/OS is a graphical shell and partially posix-compliant multitasking operating environment that runs on top of a DOS kernel." [...] "lEEt/OS is slowly but surely migrating from FreeDOS to ST-DOS, its own DOS kernel."
I don't know what it is about Finnish culture that results in such a significant number of permacomputing hackers¹ that go to great lengths to preserve old-hardware, but I am grateful for it.
¹ When I say significant I mean "at least two, but one of them is viznut"². That is significant, right?
Reading the documentation, "this is a clone of MS-DOS" turns out to be quite a misleading summary of what this is.
The documentation claims things (which I have not verified personally) such as forking, signals, and pipes. Instead of a SUBST command there is a MOUNT command. And there's a convoluted piece of hooplah that stands in place of what MS-DOS could do with the SHELL= line in CONFIG.SYS .
> Reading the documentation, "this is a clone of MS-DOS" turns out to be quite a misleading summary of what this is.
Not a "clone" in the sense of 100% compatible... but a "clone" in the sense of borrowing a significant degree of ideas and APIs/syntax from it, yet also deviating at various points (whether for good reasons or for idiosyncratic ones)
It is a clone more in the sense that the Japanese mainframe operating systems Fujitsu MSP and Hitachi VOS3 are/were clones of IBM MVS: they started out as a direct copy of MVS–due to IBM having released earlier versions into the public domain, and Fujitsu/Hitachi illegally stealing the IP of later ones–but over time diverged in incompatible directions
ST-DOS is a DOS implementation, but it is not meant to be a clone of MS-DOS. It is mostly syscall-compatible with MS-DOS, but the driver API and many other things are completely different. After all the definition of DOS is just "disk operating system".
All real mode programs that are compiled with Watcom C/C++ should work. The most recent versions of Watcom's protected mode runtime don't currently work, because they use some undocumented MS-DOS syscalls that are not implemented in ST-DOS. I intend to create a compatibility TSR that will solve most issues with those MS-DOS programs.
What's interesting is when I think about MS-DOS these days, I think I could make a functional duplicate in short order. But nobody did it at the time.
(How I'd do it is write all the semantic code in D, such as the code for EDLIN, and debug it all. Then hand-translate it into asm.)
P.S. One of the smartest things Gates/Allen did to write their original BASIC was to write an 8080 emulator on a PDP-10(?). This enabled much, much faster development than trying to hand-assemble code and toggle it in.
I'm reminded of the story of the university student who wrote some super optimised and highly convoluted implementation of some algorithm by iteratively improving it in scheme and finally translating the final scheme code to C.
I had a single game that, AFAICT, the CD-ROM version couldn't run on MS-DOS the normal way because the intro it showed needed more conventional memory that was possible with mscdex loaded[1]. There was a workaround that consisted of running the program that would resume from the latest save, then exiting to the main menu. If you had a large enough hard-drive, you could also run it by copying the data to your hard drive, then manually pointing the game at the path (thus allowing you to omit the CD-ROM driver).
I managed to get it to run off the CD by using the nwcdex (From Novell Dos 7, after Novell purchased Digital Research), which could load itself into EMS.
1: This was the CD-ROM version of Master of Magic. The intro needed approximately 630KB of conventional memory to run.
Yes. Many years ago when i worked for a computer shop, i was kinda 'forced' by the shop owner to sell it whenever a new computer was sold. He bought lots of it and nobody would buy it.
I hated it.
Back in the BBS era i loved 4DOS, but looking back in hindsight it was just a bunch of extra tools on top of MS-Dos.
Question: considering that the Mpxplay music player [1] supports native audio playback via Intel HDA on DOS (confirmed with FreeDOS), should it also work out of the box on ST-DOS?
All the HTML that worked back then still works, so I’m not sure what you’re asking? Table layouting works just as it used to. HTML 4.01 Transitional doctype works. `<body text="" bgcolor="">` works. And so on. You just write HTML. By hand. As disturbing as it may sound in 2024.
This specimen is particularly pure, not having any sort of a stylesheet (after all, we had CSS 2.0 already in 1998), the only <link> tag pointing to the RSS feed (which we didn’t have until 1999)!
Also, there was more discussion on ST-DOS (involving the author) on the Dos Ain't Dead forum: http://www.bttr-software.de/forum/board_entry.php?id=20883
FTFY
Turns out it is someone’s personal clone of MS-DOS.
Nothing to do with Atari ST, the “ST” is the author’s initials
Tl;Dr: "lEEt/OS is a graphical shell and partially posix-compliant multitasking operating environment that runs on top of a DOS kernel." [...] "lEEt/OS is slowly but surely migrating from FreeDOS to ST-DOS, its own DOS kernel."
Or STM32, as I initially suspected.
¹ When I say significant I mean "at least two, but one of them is viznut"². That is significant, right?
² http://viznut.fi/en/
The documentation claims things (which I have not verified personally) such as forking, signals, and pipes. Instead of a SUBST command there is a MOUNT command. And there's a convoluted piece of hooplah that stands in place of what MS-DOS could do with the SHELL= line in CONFIG.SYS .
Not a "clone" in the sense of 100% compatible... but a "clone" in the sense of borrowing a significant degree of ideas and APIs/syntax from it, yet also deviating at various points (whether for good reasons or for idiosyncratic ones)
It is a clone more in the sense that the Japanese mainframe operating systems Fujitsu MSP and Hitachi VOS3 are/were clones of IBM MVS: they started out as a direct copy of MVS–due to IBM having released earlier versions into the public domain, and Fujitsu/Hitachi illegally stealing the IP of later ones–but over time diverged in incompatible directions
All real mode programs that are compiled with Watcom C/C++ should work. The most recent versions of Watcom's protected mode runtime don't currently work, because they use some undocumented MS-DOS syscalls that are not implemented in ST-DOS. I intend to create a compatibility TSR that will solve most issues with those MS-DOS programs.
[1] https://github.com/microsoft/MS-DOS
> Hardware assisted multitasking was an error
Which made me chuckle.
(How I'd do it is write all the semantic code in D, such as the code for EDLIN, and debug it all. Then hand-translate it into asm.)
P.S. One of the smartest things Gates/Allen did to write their original BASIC was to write an 8080 emulator on a PDP-10(?). This enabled much, much faster development than trying to hand-assemble code and toggle it in.
My first experience was the 3.x era. Does anyone else remember DR-DOS?
I managed to get it to run off the CD by using the nwcdex (From Novell Dos 7, after Novell purchased Digital Research), which could load itself into EMS.
1: This was the CD-ROM version of Master of Magic. The intro needed approximately 630KB of conventional memory to run.
Back in the BBS era i loved 4DOS, but looking back in hindsight it was just a bunch of extra tools on top of MS-Dos.
1: https://mpxplay.sourceforge.net/
Slightly related noteworthy project: http://svardos.org/
What are the "good things from DOS"?
This specimen is particularly pure, not having any sort of a stylesheet (after all, we had CSS 2.0 already in 1998), the only <link> tag pointing to the RSS feed (which we didn’t have until 1999)!
Edit: Ooh, looks like ST-DOS is this person's MS-DOS clone. Nifty!
That was emphasised when the later 68030-based model in the series was named TT ("ThirtyTwo").
TT was then Thirty-two/Thirty-two as the '030 data bus was 32 bits.
Yours makes more sense though.