Debian Moves Closer To Voting On Proposals Over Init System Diversity

[starshipeleven]

If the result of this GR is that Debian stops support non-systemd inits, this will mean the immediate death of Devuan. No need to wait for two years then.

The sooner, the better. Actual non-systemd distros need all contributors and userbase they can get.

[ermo]

Actual non-systemd distros need all contributors and userbase they can get.

Someone linked to the *BSD-originated nosh project in the last post on this subject.

I have to admit that I wouldn't be surprised if nosh ended up being the proverbial holy grail for those who really don't like the way systemd is designed. From the front page of nosh:

It's not systemd: This is a point in its own right, for some people. A rather more sensible point of view is that it's part of maintaining heterogeneity in Linux. systemd unit files do not lock one in to systemd forevermore, because one can run them through convert-systemd-units and get nosh service bundles. And nosh is part of a family of toolsets where mix-and-match composability is a design feature.

Assuming that systemd and its units become the de facto standard for managing Linux systems going forward, it makes sense for upstream developers to ship a single set of service and configuration files targeting systemd, which keeps their testing surface as small as possible.

With nosh's ability to convert (some?) systemd files, it should be trivial to maintain a system with nosh taking on the bulk of the conversion work. In theory at least.

Nosh also has the interesting property that it is already tested on BSD. Thus, coalescing around nosh for the purposes of running Debian atop a FreeBSD kernel might be the sensible option.

Note that I have no dog in the fight per se and it's been a long time since I used a Debian-derived distribution on bare metal. I maintain a couple of packages for Solus and have found the systemd facilities useful in that respect. In the past I've also helped maintain packages that relied on sysv rc-init scripts and debugging those when something went wrong was ... well, let's just say that systemd has allowed me to focus more on the packages and less on the intricacies of service management in bash.

I also quite like the idea of the daemontools concept on which nosh is based and I hope that one of the (re-)implementations of that concept (i.e. nosh, s6, runit, daemontools-encore) rise to the top as a counterbalance to the impending systemd hegemony, because "competition improves the breed" as they say.

[pmorph]

The same thing I said above also applies to Unix and Unix phylosophies.

Actual applications care about OS APIs, and almost always there is no chance in hell that a developer can ask the OS developers to add or tweak the APIs.

Yes, that's why you avoid adding any unnecessary restrictions.

[Guest]

If they decide to ditch the idea of supporting any other init, their distro will become even worse than it already is.

[9Strike]

4. Our priorities are our users and free software

https://www.debian.org/social_contract.en.html

That has been grossly forgotten.

No they haven't forgotten it. First of all, systemd is free software and second, if the users *want* systemd, than everything is fine. On popcon they are only 45 people reporting using elogind, 44 for openrc, 520 for runit and 130 for s6. systemd on the other hand has 157516 reported users. So basically the amount of people using a systemd only system is around 99.5%, which is basically everyone.

[starshipeleven]

Init Diversity is as stupid as Desktop Diversity. A waste of effort that only brings drama.

Not really, DEs can be swapped pretty easily.

[moilami]

No they haven't forgotten it. First of all, systemd is free software and second, if the users *want* systemd, than everything is fine. On popcon they are only 45 people reporting using elogind, 44 for openrc, 520 for runit and 130 for s6. systemd on the other hand has 157516 reported users. So basically the amount of people using a systemd only system is around 99.5%, which is basically everyone.

Really? Then it is pretty much case closed unless pure overwhelming good will is offered to very minor minority.

[starshipeleven]

If they decide to ditch the idea of supporting any other init, their distro will become even worse than it already is.

Nah, they will have half-hassed systemd support the same way they have half-assed script-based-init support.

If I had a dime for all the systemd services that are in fact just wrappers to call a script or even outright scripts in the rc.d folder I still see in my Debian ARM devices... I'm so glad I migrated to OpenSUSE years ago.

[starshipeleven]

Yes, that's why you avoid adding any unnecessary restrictions.

I think I said loud and clear that you are not talking about this topic already.

[waxhead]

I think Debian should only support systemd especially since we got projects such as Devuan. For me systemd solves more problems than it creates for both desktop and server use, and the biggest issue in my point of view is that the BSD world don't have a systemd equivalent and that is where the manpower should be spent instead of discussion whether to continue to support non-systemd use cases.