@Suiseiseki @xianc78

> you can't compile it without a proprietary compiler.

I wouldn't know; DOSBox solves enough of my problems.

I attempted to figure out what it uses but they have no documentation for this in the repo, just a script that wants to run as root and six levels of indirection. (Their build system alone appears to be much larger than any version of DOS from the DOS era.) I have a subversion checkout of it from 2009 (if you believe the timestamps) and subtracting sixteen years of bloat makes it a little easier to decipher: it looks like they have written portions of FreeDOS in C++ and they want TurboC++ for it, but I can't say for certain without solving the halting problem whether it is possible to use GCC for this. (Probably not worth the effort, because it's a C++ project: it's shit.)

> I'm not too confident about the licensing situation,

MIT.

> The primary and only practical usage of "FreeDOS" is in fact the execution of old proprietary DOS software.

If you think it is difficult to port FreeDOS to use GCC instead of TurboC++, it should be easy to acknowledge that it might be difficult to port even free software written for DOS to a more recent operating system.

Free software written for DOS does exist. Some people use DOS as a bootloader! There were Linux distributions that did this, Chuck Moore did it for Forth environments, etc.

That aside, I don't see any issues with using the computer to investigate older computers and older software; I don't know why you'd object to this.