5 ms·
> - It has bugs, yes, every software has bugs. We even hit multiple of them, and some of them was fairly unpleasant and hard to find, but it can be known after
by benchaney 8y ago
> - It has bugs, yes, every software has bugs. We even hit multiple of them, and some of them was fairly unpleasant and hard to find, but it can be known after a bit of digging, because development is on GitHub. Also I'm fairly sure your shell scripts don't have any bugs in them...
Most init systems are small enough in scope that they have very few if any major issues. Systemd has them all the time. I had never had an issue with an init system other than systemd.
> - Shared knowledge. I never heard this argument, but you can jump to any other system and understand what's going on immediately if everyone uses the same init system. This is a huge advantage. Fragmenting the landscape with programs which inherently do the same thing just a bit differently is just a bad idea. I also hate split of talent and split of energy. Also, why on earth would you change to another init system ever? Or why would you ever change to an init system you don't know?
If you have an init system that is reasonable in scope, it doesn't affect your ability to change to other systems, regardless of which one they use. Most people don't need to touch their init system on a regular basis. Unless they use systemd.
> - Boot time is way better with systemd. With sysvinit, there is no parallelization. You have runlevels and that's it. I never used more than three, but I use multiple targets (own custom target also) with systemd, because it's easy and you can finally name them and debug them and see the dependency chain for every target.
This is complete nonsense. First, runit supports parallel startup, Second it isn't clear how much this matters because startup is heavily disk bound.
> - Systemd DOES NOT break backward compatibility. It can use existing sysvinit scripts. Actually it even default to those if there is both a .service file and init script.
> - "One of the critical reasons against systemd concerns the lack of portability to other Unices, and the associated risk to see Linux-based systems detach from the UNIX world."
> Anyone seeing the irony here? :D
No. This point makes no sense at all. Causing linux to drift away from other unices has nothing whatsover to do with whether or not systemd can run normal init scripts.
> I never understood systemd hate, but I guess it comes from ignorance and lack of willingness to learn about modern system design. The people who made the fork are just a bunch of hissy kids or stuck in stone age who are lazy to learn a new system IMO.
It's somewhat ironic that you are attributing bad faith to people who disagree with you when all of your points are so feeble. Also ironic is that comments like this probably contribute to how much people dislike systemd. If you have a problem with a core operating system component, that is frustrating. If you then go on to an internet forum and read about how some ignorant fool thinks that you frustration is really your fault it really pisses you off.