The debate will probably rage on until someone out there in GNU/Linux puts their foot down with all this mess and says that sysvinit is the default standard for GNU/Linux and no single OS proprietary init systems will be certified, ever, and there are no plans to remove sysvinit from this status, and all other init systems are completely optional to the distributions only.

I've tried just about all the init systems out there, but sysvinit and bsdinit are the easiest by far to use, maintain, and troubleshoot.

Bart's efforts are sound as he's working hard to have his port have the lowest possible footprint and least devastating impact.

Properly done systemd could work, but the learning curve to use it properly, as well as the fact that it becomes such a hard dependency, the developers track record, and that it's a Red Hat project call it's purpose into question, not to mention that it has forcefully deprecated several projects that were once key to all flavors of UNIX forcing many into the patchworks just to stay up-to-date and viable.

I still don't think it's viable enough to be ready. I still don't see a "need" for it. The default standards are still in favor of sysvinit, bsdinit, and even OpenRC as they're fairly much working off the same system.

I'm surprised GNU hasn't said anything due to the fact that this would also affect GNU/Hurd as well.

Tell me who and how would someone do that

Linus Torvalds?

All it would take is a community effort by a group of distributions, as well as the GNU project and Free Software Foundation itself to stake such a claim.

Stallman, for one, has remained completely and strangely silent on this matter, though in his manifesto he completely despises proprietary and closed source software, wanting complete openness with software across multiple platforms. Usually him and the FSF are all up in arms over Proprietary drivers and Closed Source drivers, firmware, software packages, etc. for GNU/Linux.

However, yes, if GNU/Linux as a whole would come together and actually forge a baseline for the operating system and place a hard label on something as being GNU/Linux certified, and not bicker over minor issues, this whole mess could be sorted out in a short amount of time. However, the only group making an decisions at this time on the future of GNU/Linux isn't even GNU, The Linux Foundation, or Free Software Foundation... it's Red Hat.

Linus also has remained silent, or maybe at this point he just doesn't care.

I have to repeat again, if you want to rant about systemd or Red Hat, this is not the appropriate thread for it. Keep this thread on topic.

4 members found this post helpful.

+1 on that.

There recently has been a revitalization of the GNU dmd project, which was originally intended for GNU/Hurd; though it should be a viable alternative for any POSIX system. In fact, current development efforts are taking place as part of the GUIX system running on GNU/Linux.

They actually restarted dmd/deco developments? That's actually good news to hear as that project was stalled for the longest time.

It's very simplistic by design and it's dependency list is very low. Not only that it's design is stand-alone and doesn't affect other programs. It's more akin to the s6 and RunIt init systems.

It's runlevels don't work right yet, and I don't think it does parallel service startup, but it could be very promising.

Speaking of which... Bart hasn't updated his finding in a while. I hope all is okay with his port.

I hope to have some small updates this week.
For now systemd-208 is stable, with some minor glitches.

for the rebuild of lots of packages,
If you don not use the session manager of pam, nothing needs to be rebuild (it can be rebuild)

as the packages stands now, everything runs stable, and got no crashes over the last 3 months on 2 systems.
1 which is having a reboot every 2-3 days.
the other running constantly.

Currently I am in the process of moving countries for my work, so can spend little time on this.

So Linus should say something and stop with it? I don't like systemd either but isn't this free software after all?

2 members found this post helpful.

We have the freedom to fork any problematical package, or to write a better equivalent from scratch. This is regularly pointed out by the self appointed plumbers.

Of course, it's almost impossible for individual volunteers to come together and successfully fork or reimplement a constantly morphing ecosystem that is created and controlled by full-time employees of a billion dollar company. They know that; we know that. The Linux Foundation seems to be entirely comfortable with that. Well, they would be, wouldn't they. Linus is only going to get assertive when the kernel is technically impacted (like the time he yelled at Kay).

And so here we are. We've become the counterculture of 21st century Linux. Let's embrace that, instead of looking for some authority figure to make us mainstream.

1 members found this post helpful.

I am at this point really fed up with this. Not with systemd or anything in that regard, but with some LQ members participating in this thread. There were already two moderator warnings to keep this thread on-topic. The topic of this thread is bartgymnast's effort to port systemd to Slackware and the problems he faces with that, together with the progress he makes.
Anything else, including the formatting of the thread title, Red Hat, general direction of systemd and Linux, Linus Torvalds, the FSF and Richard Stallman, the process of creating standards, free software in general, ..., are not topic of this thread.
If you want to talk about that do it in a different thread.

I will not tolerate further derailment of this thread.

6 members found this post helpful.

^H^H^H^H

I finally have some time to test this, but I had some problems compiling systemd, it bailed out with different error messages every time. I had to force compiling with -j1.
Does anyone else have that problem?

What were the error messages?