Systemd 249-rc1 Released With Many New Features

[jacob]

Your book and wiki article are both written after-the-fact (book for example 5 years after initial release). Go try and find arguments and discussions revolving around systemd when it wasn't adapted yet. Especially in various forums.
I recall proponents yelling about the need for "better init" for the most part, bit less about the need for unified system services manager tho that came up and was mentioned sometimes as well.
But when it came to back- and forth arguments, majority of the whine and justification was revolving around "the need for better init" even tho Linux at the time had access to multiple alternatives.

Literally none of the main developers-proponents made squeak about planned massive scope- and feature creep or it might have scared off too many people and made adaption harder.

In addition to what others have said I would point out that very quickly, Lennart and Kay were joined by many developers, some of whom brought in projects that were completely separate up to that point. So not only there was never any secret conspiracy involving systemd, but the community at large was never "scared" of it, rather they thought it was promising, joined in in droves (over 300 active committers as of now) and embraced it precisely to implement many new tools and features.

The "multiple alternatives" chorus is a scratched record at this point. Just to repeat what had been said ad-nauseam, no-one gave a f**k about "init alternatives" until systemd came about, no major distro ever offered "init alternatives" (having openrc or daemontools in a contrib repo is not the same as actually basing the OS on it), those who whine about systemd always claim how ***BSD is supposedly better but there is absolutely no alternative to anything on BSD, many of them happily use MacOS which has launchd and no alternatives, there are no alternatives for the TCP/IP stack, the syscall interface, GCC (until clang came around) and it doesn't bother anyone, there have been some alternative/competing projects at various times (wayland vs. mir, before that Xfree vs X.org, rust coreutils vs gnu coreutils etc., snap vs appimage vs flatpak) and each time the same crowd who is whining because lack of "init alternatives" was howling about fragmentation. So there.

[aht0]

In addition to what others have said I would point out that very quickly, Lennart and Kay were joined by many developers, some of whom brought in projects that were completely separate up to that point. So not only there was never any secret conspiracy involving systemd, but the community at large was never "scared" of it, rather they thought it was promising, joined in in droves (over 300 active committers as of now) and embraced it precisely to implement many new tools and features.

The "multiple alternatives" chorus is a scratched record at this point. Just to repeat what had been said ad-nauseam, no-one gave a f**k about "init alternatives" until systemd came about, no major distro ever offered "init alternatives" (having openrc or daemontools in a contrib repo is not the same as actually basing the OS on it), those who whine about systemd always claim how ***BSD is supposedly better but there is absolutely no alternative to anything on BSD, many of them happily use MacOS which has launchd and no alternatives, there are no alternatives for the TCP/IP stack, the syscall interface, GCC (until clang came around) and it doesn't bother anyone, there have been some alternative/competing projects at various times (wayland vs. mir, before that Xfree vs X.org, rust coreutils vs gnu coreutils etc., snap vs appimage vs flatpak) and each time the same crowd who is whining because lack of "init alternatives" was howling about fragmentation. So there.

If "everybody were happy about systemd" there wouldn't have been massive flame wars across the effin internet (not just moronix).. Nor you guys would feel the need to jump into every systemd theread in Phoronix and start proactively bashing all potential dissenters before any has even shown up. Or jump into every BSD thread to announce your opinion about linux supremacy (doers of both deeds generally overlap to 90% in phoronix btw).

Don't you even detect cognitive dissonance in your claims? If nobody gave fuck about potential init alternatives then why the fuck you folks whined so much about systemv problems and justified systemd in particular as "badly needed init alternative".
It just had to be pushed down distributors throats nomatter what, to get it adopted the degree RHEL wanted so they could do whatever they wanted to do with it but didn't dare yet come out publicly.

Those " your many developers" are mostly RHEL or some other corporate paid (Facebook for example) devs.. 5 devs have over 1000 commits while Poeterring is leading the pack by wide margin. When you set threshold at 50 commits you'll also get less than 50 devs having done contributing work. Rest of the 1500+ devs haven't ... nah, not worth bringing them out even.

It's pretty much RHEL and corporate doing whatever the hell they want when it comes to your "now-crucial OS component" and your say is pretty much limited to swallowing it whole or choking on it.. If you dont like it, you have no good alternative. Bitchboys for the big businesses.

[jacob]

If "everybody were happy about systemd" there wouldn't have been massive flame wars across the effin internet (not just moronix).. Nor you guys would feel the need to jump into every systemd theread in Phoronix and start proactively bashing all potential dissenters before any has even shown up. Or jump into every BSD thread to announce your opinion about linux supremacy (doers of both deeds generally overlap to 90% in phoronix btw).

Don't you even detect cognitive dissonance in your claims? If nobody gave fuck about potential init alternatives then why the fuck you folks whined so much about systemv problems and justified systemd in particular as "badly needed init alternative".
It just had to be pushed down distributors throats nomatter what, to get it adopted the degree RHEL wanted so they could do whatever they wanted to do with it but didn't dare yet come out publicly.

Those " your many developers" are mostly RHEL or some other corporate paid (Facebook for example) devs.. 5 devs have over 1000 commits while Poeterring is leading the pack by wide margin. When you set threshold at 50 commits you'll also get less than 50 devs having done contributing work. Rest of the 1500+ devs haven't ... nah, not worth bringing them out even.

It's pretty much RHEL and corporate doing whatever the hell they want when it comes to your "now-crucial OS component" and your say is pretty much limited to swallowing it whole or choking on it.. If you dont like it, you have no good alternative. Bitchboys for the big businesses.

The "massive flamewars" were from a hypervocal but extremely tiny minority. As always the proof is in the pudding, look at the combined user base of Devuan, Obarun, Void, Slackware & co compared to even one single mainstream (i.e. systemd-based) distro and you will have your answer. And yes, most (but not all) of the devs are Red Hat and other corporates, just like, you know, EVERYTHING. Linux (the kernel), GCC, LLVM, Mesa, GNOME, even KDE are all developed mainly by commercial organizations and I'm very happy with that, because the alternative is that these same orgs would develop proprietary software instead and FOSS would be reduced to a spare time toy that no-one takes seriously for real work. In fact when RMS started the GNU project his first step was to court the IBMs and other companies that operated at the time for involvement, sponsorship or funding. He wasn't successful because they all thought his idea was some hippie BS. But as the same RMS famously said, the goal was never to stop people making money with software, it was to make it possible for people to make money with Free software. On the other hand, some people never seemed to understand what FOSS means: it means simply if you are not happy, you are free to patch or fork and make it work your way. It has never meant that special snowflakes would be entitled to have a say in how other people's projects are ran, it also never meant that it would be mandatory to stick to POSIX or be written in shell scripts or forever imitate the design and workflow of the proprietary OS called Unix.

Systemd itself is entirely free and open source, unlike your beloved SysV which has always been strictly proprietary. But for what must be the 1837864846th time, it was NEVER meant to be an "init alternative". From Poettering's first announce it was meant to be an USERLAND BASE SYSTEM. Not one that fits well within the traditional *nix environment, but one that REPLACES IT. That's what made its success, distro maintainers and software developers saw an opportunity to finally get rid of the stickytape, DYI, mostly-kind-of-works-for-me scripts and piecemeal hacks to work around problems that shouldn't have existed in the first place, and replace it all with an actually designed and planned solution that takes into account modern hardware, modern user requirements and modern workflows rather than expecting users to bend backwards to content themselves with 1980s technology. This is also the reason why no-one except a few VUAs and self-appointed ideological komissars really gives a cr*p about the "alternatives", whether they are called runit, openrc, daemontools, s6 or whatever. These are all pure init replacements, whereas systemd offers to move away from the notion of "init" in the *nix sense altogether. It has been massively embraced precisely because of that, not despite that as you seem to believe.