Debian Moves Closer To Voting On Proposals Over Init System Diversity

[mrugiero]

You're missing the point. It's not about consumer preference, but the ability to customize system software to suit a particular set of needs. As you'll know, Linux is being used in a vast array of devices and environments, to suit applications with an even greater set of needs and requirements. To think that one userspace blob of services can adequately satisfy them all is beyond hubris - it's insanity.

To think all of them would use Debian and big desktops (the only ones depending explicitly on systemd facilities that I can recall) is insanity.

Maybe not for you, what one thing I always found appealing was its openness and flexibility. I think those have been virtues, and I hate to see ever-increasing parts of the userspace get absorbed by the sytemd borg (maybe that analogy will make more sense to certain individuals of a more interstellar predisposition), eroding away at some of that flexibility as is does.

They may be virtues in theory, but fragmentation is a thing, and it wastes man power, of which we have way less than we need to get anything on-par with closed alternatives.
It also wastes admins time when they have to deal with several disparate distributions that choose to do things differently of each other just because.

I just wonder how long it'll be, before systemd starts to infect the kernel.

If we count bus1, right now. But they don't have free reign there, proved time and again by getting stuff rejected loudly.

[coder]

To think all of them would use Debian and big desktops (the only ones depending explicitly on systemd facilities that I can recall) is insanity.

They may be virtues in theory, but fragmentation is a thing, and it wastes man power, of which we have way less than we need to get anything on-par with closed alternatives.
It also wastes admins time when they have to deal with several disparate distributions that choose to do things differently of each other just because.

The point about unsupported alternatives and lack of manpower actually go together, as well. If systemd absorbs all the oxygen, then the alternatives stagnate and systemd becomes the only way.

The vision you're putting forth is basically that Linux should be an open source Windows. Windows has different userspace services, and one could theoretically replace one with a custom implementation, but people only run the configuration that MS gives them. That might be fine for most people, most of the time, but when you want to push Windows outside its comfort zone, the lack of options can really be problematic.

[mrugiero]

The point about unsupported alternatives and lack of manpower actually go together, as well. If systemd absorbs all the oxygen, then the alternatives stagnate and systemd becomes the only way.

So? Manpower is distributed around projects. More active projects == less manpower/project. I'd rather have one high quality product than a hundred barely usable (if lucky) ones.

The vision you're putting forth is basically that Linux should be an open source Windows. Windows has different userspace services, and one could theoretically replace one with a custom implementation, but people only run the configuration that MS gives them. That might be fine for most people, most of the time, but when you want to push Windows outside its comfort zone, the lack of options can really be problematic.

The vision I'm putting forth is basically not fragment for nothing, and not expect everyone to work for the few who want to "push Linux out of its comfort zone".
Having multiple inits in a single distro is, in my eyes, part of fragmenting for nothing.
Not everyone has to use Debian. Most of the actual valid use-cases not covered by systemd aren't well covered by Debian either way.
You wouldn't use Debian on a router. You'd use something like OpenWRT, Buildroot or Yocto.
You would use Debian mostly on desktops (where systemd is appropriate) or on servers (where systemd is also appropriate).
systemd was also used successfully in some embedded environments, as shown here: https://github.com/profusion/demysti...bedded-systems
That said, I'm OK with people pushing patches to scratch their itches, and taking up the task to maintain those, as long as that doesn't degrade everyone else's experience.
I just think they're wasting their time, but it's theirs to waste.

[coder]

So? Manpower is distributed around projects. More active projects == less manpower/project. I'd rather have one high quality product than a hundred barely usable (if lucky) ones.

Again, I think the MS Windows analogy in instructive, here. The lack of alternatives doesn't mean your only viable option is necessarily satisfactory for all needs and purposes.

Having multiple inits in a single distro is, in my eyes, part of fragmenting for nothing.

Heh, you've fallen into the conceptual trap of thinking about systemd as an init system. It's not. It's a growing constellation of userspace services, with service management near the gravitational center. As long as you regard it as an init system, you're missing the point.

I don't actually care about a distro supporting only one init system. It sounds like an entirely pragmatic choice. I simply choose to believe that it shouldn't lock you out of other choices about your userspace services, which is the practical reality we face.