> This is a legitimate question, since I am not a Windows user: has it really? Or are you referring to GNU software[0], which is not what we are referring to when we say GNU is a fully free Unix replacement; GNU software is only a part of that.[1]

As far as I can tell this subsystem does nothing that cygwin hasn't done in terms of 'bringing gnu to windows' since 1995, except that it does it technically better and allows for you to use software that does not depend on already ported components without recompilation. This means, for example, that you can use a libc other than glibc without porting that libc to windows first (ie. you could use musl, which is absolutely not part of the gnu system and even implements its own dynamic linker). You couldn't do that with cygwin. At least not trivially.

> Sure they have. But that's not what the conversation is focused on: it's focused on being able to use GNU Bash and all the other Unix (mostly GNU) utilities that hackers are used to using.

Even if I concede that GNU constitutes a 'system' in the way that RMS insists (which I don't, at all. His claims are political, not technical, imo), I would not agree that using this layer constitutes using that system for precisely this reason. Using bash or even gcc is not sufficient to consider a system 'gnu', even by RMS' own statements, else OSX and the other BSD systems would be gnu systems as well. For the most part, use of "bash on windows" is literally just that. You run bash and it exists in a space that is like but not exactly the same as linux, but that part is largely artifice. You don't even get a complete /proc filesystem.

[Note: Considering I am literally referring to those RMS screeds in my posts, I'd appreciate it if you at least pretended to believe I've read them.]