Re: This is perfect for a Friday story
There are very rational reasons why many people have valid reservations about Lennart Lennart Poettering's misguided creation and Nala Ginrut's observations are presented below:
"Few years ago, I was working for SUSE Linux Enterprise (SLE) as full-time developer. My boss asked me if I’m interested in maintaining systemd for SLE. At that time I know little about systemd. Then I said “I’ll see what I can do”.
I’ve downloaded the source code accompanied with 2000+ backport patches, this took almost a half day. When I was waiting, I had reviewed the code and try to find out what’s in it. And I had contacted several experienced colleagues to learn about it from them.
Finally, I could figure out that systemd takes advantages of Linux kernel things to start services in parallel. This sounds good, but it introduces unbelievable complexity for such a functionality. If you’re a common desktop user, maybe you rarely encounter problems, or you may endure some bugs in systemd. But for an enterprise version Linux, it makes the work of maintainer hard.
And the 2000+ backport patches implies that I (as the maintainer) have to maintain all of them alone, since it’s backported. I discussed with my boss, and told him this kind of work requires a small group people, rather than one man effort. Fortunately, my boss agreed with me completely, so he managed to throw out this package to a bigger team.
Even now, I still remember the complexity in systemd code, and I always tell my engineer team to avoid such kind of complexity in the project. Eliminating complexity is far more important than adding features without clear mind. Don’t try to put every good features into just one project. Every feature is good, but package all of them, you’ll get shit.
I don’t know who got this package finally, but I hope she/he is fine…"
^ The home user might not notice systemd's effects and consequences but for those of us who work in deployed enterprise Linux situations, we do get to see the downsides of systemd.