[md8z]

To me that just further illustrates the point that GNU failed. I mean it's right there in the name, it's supposed to be "GNU's not Unix" but the only parts anyone actually uses are clones of Unix that get used on a different Unix-like operating system (Linux) which is not GNU. And the desktop environments also don't have anything to do with GNU. IIRC the only official desktop environment of GNU is GNUstep, which I don't even think is actually packaged by any distros at the moment.

[kevin_thibedeau]

GNU's not SysV/BSD UNIX. It isn't literally meant to be for a different OS altogether. Even Hurd has a POSIX syscall API.

[md8z]

Come on now, that is a pretty big stretch. And it's not even true, coreutils and glibc implement a lot of SysV and BSD compatibility. My point here is that GNU as an operating system is dead. People can make up some new goals for it but it seems obvious that the original goals have mostly been a failure. Even if you disregard Hurd and count the "FSF approved" Linux distributions, few people actually use those because most of them are just ordinary Linux distributions but with the only major change being that some proprietary packages and drivers are removed. Is that really adding value to the software ecosystem? I personally wouldn't recommend those to anyone beyond hardcore GNU nerds, for most people I'd still say just use Ubuntu.

[GekkePrutser]

The desktop environments are not from GNU, no. But they are all GPL. In that sense the FSF still has a strong influence IMO.

[md8z]

Yes the desktop environments are still using GPL but those are already established projects, not growth areas. For newer projects it seems there is a shift away from copyleft towards Apache-style licensing. The newer copyleft licenses (GPL3, AGPL) just haven't gotten that good of a reception. And in the areas I've seen that do use them it seems it's despite the FSF, not because of them, for example the FSF seems to have completely given up on license enforcement or helping with anything in that area.