86-DOS running on a PC

[xelloss]

Michal Necasek has managed to put together a version of 86-DOS 1.0 that will run on an (IBM or compatible) PC.
It uses the pre-release PC-DOS IBMBIO along with some small patches to the 86-DOS kernel.

Remember that PC-DOS 1.0 is an OEM release of 86-DOS 1.14, so this is the most ancient DOS version to ever run on a PC.

https://www.os2museum.com/wp/pc-86-dos/

Enjoy!

[ToMi1]

That's interesting indeed. Is it able to run on an emulator like 86Box or PCem?

[xelloss]

That's interesting indeed. Is it able to run on an emulator like 86Box or PCem?

I haven't checked myself yet, but it should.

[tarquin44]

Yes, it runs in 86Box.

[daarch]

Quite an achievement.

In case someone else wants to try it under Qemu, the image must be patched to include the 0xAA55 signature at the end of the boot sector to be usable there.

[Iforgor]

I was always curious if Tim Patterson really ripped off CP/M. QDOS was suspiciously similar from UX point of view, but was it ever investigated from binary code perspective? I believe DR sold the source code to all the interested parties at that time, so it's not rocket science to violate copyright, if a person has coding background, which Tim had back then.

[vbdasc]

I was always curious if Tim Patterson really ripped off CP/M. QDOS was suspiciously similar from UX point of view, but was it ever investigated from binary code perspective? I believe DR sold the source code to all the interested parties at that time, so it's not rocket science to violate copyright, if a person has coding background, which Tim had back then.

I am personally almost sure than no copyright infringement was made in this case. Yes, QDOS was made to look superficially similar to CP/M. But this means nothing. The most important thing is that at this time, CP/M was mainly an Intel 8080/8085/Zilog Z80 thing, while QDOS was an Intel 8086 code, and there is no binary compatibility between these CPUs. I can assure you that dumb recompiling of i8080 code to i8086 for a project like a DOS won't yield anything but bloated and buggy code, so it's not worth the effort.

There is also another thing. Both CP/M and QDOS have very small sizes. They're both totally feasible to disassemble and analyze. I just can't imagine that any of the thousands of competent MS-haters couldn't ever come with a proof for the ripping off. They had 40 years for that, for Pete's sake.