Can you explain the use of the word modern? Is it written using a modern language? It doesn't look like any of the ideas presented by it are particularly recent. In any case, of course we have it in ports for you to use.How about modern faster init system as runit or whatever?
Why do you think it needs to change?Should it be changed or it's decent?
gkrellm &
urxvt &
xfe &
fluxbox start
Because of FreeBSD's design, it is "relatively" easy to port another init system to be used instead of the standard rc. (Hopefully it's POSIX etc)Hello. What do you think about current initialisation system in freebsd? Should it be changed or it's decent?
How about modern faster init system as runit or whatever?
I hope there is a port of SysV-init used by Devuan so I could reuse init scripts from Devuan. I tried to use Devuan's init scripts with our RC, it will not work. The syntax is incompatible.Because of FreeBSD's design, it is "relatively" easy to port another init system to be used instead of the standard rc. (Hopefully it's POSIX etc)
GhostBSD has done it, for example. There are others available in ports, such as runit. Go for it, try it and let us know.
So "SystemD" is the forbidden word? Perhaps SirDice's "Why FreeBSD not more like..." is sufficient to answer, I don't need to write too explicit like this.Well, that didn't take long for the thing that shall not be mentioned to be mentioned...
But that's not the reason, SysVinit works fine with any POSIX shell as long as the actual init-scripts don't have bashisms. Debian has used dash for /bin/sh for a long time.And obnoxious, Linux has bash for such things, we use sh.
Why is that the current "buzz phrase" for so many answers to so many questions?It depends on the use case.
You could be more vague but those sound like applications, not operating system issues.it does lack in some areas such as parallelism, on-demand start/stop, recovery options, and more complex dependency management.
It depends on the use case.
I love the current system for it's simplicity, but it does lack in some areas such as parallelism, on-demand start/stop, recovery options, and more complex dependency management. It doesn't make sense to compare rc.d() with systemd, the two solve different problems.
We've been here before. This is a road well travelled.The idea behind systemd (even for all the negativity around it) is actually very good. I think the hate around it was due to the way it was "sold" as an init system, but instead delivered a lot more. It's something completely new that had never existed before (at least in Unix/Linux). I like the concept of a "system layer" that sits between kernel and userland. Such a design works very well for "modern" desktop/laptop platforms. I personally wouldn't mind seeing such a system layer appear in FreeBSD, maybe even something that cooperates with rc.d() instead of replacing it. In a way, devd() is kind of like this.
For servers I wouldn't change a thing, rc.d() works perfectly. Servers already take several minutes to just run through the POST process. An extra delay with booting up the OS and starting all the services in sequence doesn't add much to an already long wait. Additionally, you don't really need all that extra fancy stuff systemd brings running on a server.
Yes. I lost failure's password. I have bad luck with Windows 10 so I'm here again, only to struggle with compiling IUP on FreeBSD. My writing style never changed and I have no need to hide it.
Nah, OP doesn't match the style. Let me rephrase: obnoxious == gh_origin.Oh you reckon boombim == obnoxious == failure == gh_*?