Systemd fallout: Debian fork Devuan set up

The problems with Upstart weren't about whether it was "unix-y", but about Canonical's licensing requirements. I groused a bit to myself about having to learn a new init system -- but when I looked into it the logic was sound, it made sense, and offered clear benefits, so I accepted it. (And funnily enough, those license requirements were what hobbled it into a slow decline and inability to compete effectively against systemd).

The problems with Mir and Wayland were about whether they were going to be unix-y in certain respects, and some nasty gamesmanship between the competing projects (I'm not going to get specific, but one side was mostly responsible for that). As it became clear that some important principles would be observed and certain functionality would remain accessible, the debate died down into a "we'll keep an eye on it, but it looks like it's going to be OK (Wayland, at least -- I'm not sure any more about Mir).

Systemd is a different kind of a mess -- it's seen as a fundamentally bad design decision, that even worse, insinuates itself into components and processes outside its proper scope, in a way that make the OS needlessly complex (under the guise of cleaning up complexity, yet) more fragile, harder to troubleshoot, etc, etc. It's like all those "user-friendly" features we've seen elsewhere, that supposedly make life easier for the user, but end up being a major source of problems, themselves.

On top of that, the interlocking nature of some dev programs / projects (eg. most obviously, Gnome) is resulting in the effective binding of large swathes of supposedly independent code and components to systemd, that makes it arbitrarily difficult for those who want to stay clear of this design framework, to actually do so. (In sites and forums for various distros, there have been many admissions that the distro is "switching" to systemd, simply because the effort (time and man-power) makes it impractical to do anything else, regardless of whether they think it's a good direction.)

Devuan is a concrete demonstration of how different this mess is from earlier "tempests in teapots". Experienced long-timers aren't just complaining and bitching; they're actually trying to fork one of the very biggest FOSS projects out there -- this is bigger than when xfree86 forked to Xorg.

Quoting:Systemd is a different kind of a mess -- it's seen as a fundamentally bad design decision, that even worse, insinuates itself into components and processes outside its proper scope, in a way that make the OS needlessly complex (under the guise of cleaning up complexity, yet) more fragile, harder to troubleshoot, etc, etc. It's like all those "user-friendly" features we've seen elsewhere, that supposedly make life easier for the user, but end up being a major source of problems, themselves.

On top of that, the interlocking nature of some dev programs / projects (eg. most obviously, Gnome) is resulting in the effective binding of large swathes of supposedly independent code and components to systemd, that makes it arbitrarily difficult for those who want to stay clear of this design framework, to actually do so. (In sites and forums for various distros, there have been many admissions that the distro is "switching" to systemd, simply because the effort (time and man-power) makes it impractical to do anything else, regardless of whether they think it's a good direction.)

Devuan is a concrete demonstration of how different this mess is from earlier "tempests in teapots". Experienced long-timers aren't just complaining and bitching; they're actually trying to fork one of the very biggest FOSS projects out there -- this is bigger than when xfree86 forked to Xorg.

Perhaps this Devuan forking possibility is also an immensely viable response to repeated calls for action by the very systemd advocates themselves; calls for action such as those at 'Don't boycott systemd', found at http://dontboycottsystemd.org/

Quoting:Do something productive

Stop talking. Or stop talking and do something. Or just talk and do something. Give a solution or contribute to a solution of your problem. Pay much more attention and effort to a solution than to a problem.