PC-MOS/386 Source Code

I missed this when it was initially announced. The source code for PC-MOS/386 version 5.01 is now available on github under the GPLv3 license. It requires the user to supply Borland C++ 3.1 in order to build, but there are binaries checked in as well, including a bootable floppy image.

PC-MOS is a multi-tasking/multi-user DOS clone. It was one of the first commercial products which used the 386’s virtual-8086 mode when it was released in early 1987 (but not the first, that was almost certainly CEMM in 1986).

It is worth noting that PC-MOS also supported 286 systems. It required custom hardware and also used the well-undocumented LOADALL instruction (see MOXMEM.INC).

The source code is unfortunately not organized in a very clear fashion; there are for example three separate and not quite identical copies of the PC-MOS kernel.

There is an interesting reference to “Bimini 2-SX” aka IBM 486SLC, which apparently had a bug affecting PC-MOS/386: “When executing a CALL or JMP instruction with a negative
displacement within a 32-bit code segment and an operand prefix override is used (i.e., to limit the target address to 16 bits), the resulting target address is computed incorrectly.” (see _386.ASM in the mos5src directory for additional detail).

This entry was posted in 386, DOS, Source code. Bookmark the permalink.

8 Responses to PC-MOS/386 Source Code

  1. Cassiel andersson says:

    Hi is this something for you?

    Citrix win view för networks v2.3
    Microsoft powerpoint för windows 1990-91
    Lotus notes server för OS/2
    Lotus notes Arbets station för windows v3.0c
    Lotus Notes windows v3.0c
    Lotus notes server OS/2 v3.0c
    Lotus notes OS/2 v3.0
    Microsoft Axel för windows v5.0
    Lotus notes data disk
    Lotus Notes Release 3 client för windows
    Lotus notes release 3 mediaclips
    Lotus notes release 3 windows program international edition
    AU system Avitel windows svensk v2.11
    Lotus notes sample application
    Lotus notes Release 3 OS/2 program
    Lotus notes v3.0c
    Microsoft windows v3.1 HP swedish localisation
    Central point pc tools för windows
    Lotus freelance graphicks v2.01 för win
    Lotus inprove 2.0 för windows
    Tulop computers win TM 3.1
    Discserv v3.11x
    Lotus organizer evaluation copy orginal disk
    Netware NML Library update disk Ink LIBUP3.exe
    Central point PC tools sv dos shell&file viewers
    Olicom Token ring driver disk
    Wordperfect presentation v2.0
    Lotus organizer 1.0
    Lotus Datalens för Lotus notes för windows v3.0
    Corel SCSI network manager netware modules v1.0
    Central point backup för windows v7.2
    Microsoft Exel programdiskett
    IBM OS/” Warp v3.0
    Recognita plus international v1.1w release
    Adobe typ manager win ver

  2. Julien Oster says:

    Neat. I’m looking forward to see how that extra piece of hardware (apparently a “CHARGE” card) worked. There is also mention of an “ALL CARD (FOR 8088)”, which could potentially be even more interesting, given that the 8088 of course had no MMU at all to speak of:

    I386 equ FALSE ; IF INTEL 80386 CPU

    But maybe that comment is misleading. I also wonder why it says 8088 instead of 8086. Was this actually something specific to the 8088?

    Guess we’ll find out.

  3. Michal Necasek says:

    That’s probably more something for archive.org. Mostly in Swedish I assume?

  4. Michal Necasek says:

    What I read somewhere is that the PC/MOS custom 286 hardware included some kind of interposer between the CPU and motherboard. So if they had something similar for 8086 class hardware, the interposer would have to be different for 8088 and 8086. And all the early PCs used 8088 CPUs. IBM at least didn’t offer 8086-based systems until 1987 I believe, years after 286s. So if there was something, it may well have only supported 8088 and not 8086 CPUs.

  5. Christian says:

    @Cassiel: I’d be extremely interested in WinView for Networks. Can I contact you using email with regards to it?

  6. Richard Wells says:

    The Eagle Turbo XL from 1984* was an XT clone with an 8086 processor; it may have been the first PC compatible with 8086. By that time, the All Card was being phased out. Moving the video addresses to increase contiguous memory required patching applications but All Computers only managed to make a small number of patches and worse still did not work at all if the system had EGA. Being able to turn part of the 640k into expanded memory wasn’t much benefit.

    The All Charge Card had similar problems but at least extended memory could be turned into expanded. Worked wonderfully with the serial port text only Multilink** which made the AT Gizmo such a nice idea. Been a long time but I thought AT Gizmo was a licensed version of Charge Card. Ran into a bit of a problem once 386 prices started to decline; no one would pay $300 to give a 286 system a limited portion of 386 capabilities.

    * Well, 1983’s Eagle had a copyright issue with its original ROMs. Not sure if any were sold before IBM legal noticed.

    ** Multilink*** was the precursor to PC-MOS that ran on top of some versions of MS-DOS.

    *** Multilink had combined with the Tall Tree JRAM memory card to run background processes on an XT speedily making it one of the few programs that used (or even understood) the Tall Tree memory standard. Proved the concept of large DOS sessions which set the stage for newer more general purpose solutions.

  7. Bill S says:

    If I recall correctly AT&T / Olivetti had some extra hardware on the PC 6300+ with a 286 that would intercept signals on the bus and give you some of the features that would later exist in the 386’s v8086 mode. I own one of the machines, I’ve meant to look into it more but never got around to it. The idea was that System V could run in 286 protected mode and the hardware would handle switching modes and virtualizing certain machine states, etc for DOS. I’d imagine its similar in concept.

  8. ForOldHack says:

    The AT&T 6300. and +. Wikipedia has some good information on that: the 6300 ran Venix/86 and the 6300+ ran DOS and Sys V III under Locus’s Dos Merge, Simultask 1.0 was based on Locus Merge, Simultask 2.0 was based on VP/IX, and then the product became Univel UnixWare 1.0 Personal Edition came with DOS Merge 3.0 and Novell’s DR DOS.


    “The 6300 Plus was designed with special hardware on the bus that would suppress and capture bus cycles from the DOS program if they were directed toward addresses not assigned for direct access by the DOS virtual machine.”

    https://en.wikipedia.org/wiki/Merge_(software) Special hardware confirmed and for what purpose.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.