The freely available MSC 5.1 that ships with the freely available MSDOS 4.01 (I think that's the version), is good enough to compile the PDOS source code, to produce either PDOS/86 or as the io.sys component of the 32-bit PDOS/386 which is compiled with gcc 3.2.3 normally.
I was able to find a technique that allowed the free MASM 5.1 to interface with the 32-bit assembler. A lucky breakthrough really.
MSC 5.1 has huge memory model support which is what my design requires.
So I think this means that historically, around 1988, I could have built 16-bit IO.SYS and other components (ie MBR + VBR) on an XT using Microsoft C and MASM, and then on an Amiga I could have built gcc as a cross-compiler to 80386 to produce pdos.sys (equivalent of msdos.sys) and command.exe (equivalent of command.com) and been able to stand up a bootable 80386 OS.
That's "if I knew then what I know now", but also, "if the software had been written a hell of a lot quicker".
Fast forward to 2026 and the practical implications are that I am now in a position to ship the Microsoft tools with Freewindows and I don't need Watcom with its more complicated license. I don't know if this is of any practical use though. But my work exists as more of a backstop than of any practical use. |
Thread view
Board view
