I like systemd as a way to manage services. I don't like how they sprawl of it, as if everything is a new problem that has never been solved before. I don't like how disto maintainers are just eating up these half-baked features which make them difficult to turn off.<p>Why does systemd need it's own logging magic? I guess logging was never done prior to systemd. They literally had one corner case to solve for which means they decided to reinvent logging around instead of just solving for that one issue in already mature solutions.<p>Why does it need to own time sync... Because I guess they feel that can do it in a more complete, more secure, and less buggy way than the existing mature products.<p>I could go on and on about this nonsense that has been shoved down the users throat by the enterprise Linux giants. The sad thing is, there now seems to be a whole crop of users who don't remember the core philosophy of do one thing and do it very very well. So there's been an explosion of other things that try and suck in everything and duplicate efforts. They all seem to do their core problem well but all the ancillary things that they reinvent along the way they totally suck at.<p>Bringing up freeBSD. I'll deal with the irritating pain of systemd and the Enterprise Linux overlords then deal with the whole host of problems that come with FreeBSD
This sounds a bit like the question "Have you stopped beating your wife?" (perhaps today better to say spouse or SO). You're damned whatever you answer.
Having said that, I think systemd can be an excellent tool suite. The thing is that often in lagacy production environments (which much of the world runs on) one must use what is available, and built in toola like cron are always there.
This is just my own personal preference but I have moved away from distributions that use systemd for my own things. I like the idea of systemd. I do not like the implementation. It has spread from being a replacement for init to being its own OS framework, logging, optional resolver that I assume one day will not be optional, network and namespace integration and so much more. The hooks into dbus and binfmt_misc are laying the groundwork for bringing back windows 95 design vulnerabilities. So rather than arguing about it I just moved on.