This is a fair comment. RedHat's implementation of SysVinit was quite different to Debian's... and both were substantially less palatable to me than Slackware's init, which is often (improperly, IMO) described as BSD-style.

None of them were as simple as AUTOEXEC.BAT, and with good reason.
For me, as a "techy hobbyist," type person, I like to see all of the details. Whenever I'm forced to use Windows, I am often sitting, staring at the HDD light blinking like crazy wondering what the heck it's doing... and there is real no way to find out, or the answers it gives you are bogus.

systemd, being based upon launchd (an Apple product), does seem to hide things... or at least tries to make them less visible, because the user doesn't need to know. Well, no maybe I don't need to know, but I want to know.

The important thing is that systemd remains a modular/replaceable piece, and that it's use is not forced upon anyone. That is what is critical here, and I hope beyond hope that the Debian developers recognise the importance of maintaining this freedom.

Oh, which part is exactly hidden (and how well did they hide that)?
I see no connection. Clearly, systemd is a modular software (regardless of debian) and will remain so in the future. The Debian developers cannot, does not want and will not change this freedom in any way (regardless of systemd)

Here's a recent example of what I was saying: https://unix.stackexchange.com/quest...unit-lifecycle

"So, after firstboot, whenever i would start this machine with a host without access to the ntp server, systemd will just gaslight me.

systemctl start sshd <- nothing happens. no failure. no timeouts. nothing on any log! Wait for 3h and still no idea what is happening."
...
"I enable network and try a naive restart systemd-timedated. Of course, it fails with absolutely no information about why. not even what it was trying to do. just fail."

A following UP keystroke followed by systemctl cat sshd should lead to discovery what was expected to happen. systemctl edit sshd instead would embrace tweaking the expectation. There's no need to hunt for an explanation, or where to go to fix it if it seems to be broken.

I still remember my first time using pre-takeover RedHat Linux which I got by purchasing a manual that looked like a copy of your local phone companies' white and yellow pages combined. My baptism was one by fire because then I was forced to learn how to properly administer a Linux system using the old school SysV init scripts. I can still remember that in that period almost every distribution that I could find was using some implementation of SysV init or another. Then when the shift to SystemD came around I was rather upset because I had to relearn everything that I already knew from SysV init.

It was through this experience that I figured out that I can tolerate pretty much whatever stupid ass thing that these corporations do. I do believe that the pre-IBM(BigBlue) buyout was already throughgoing quite a bunch of dough optimizing the boot process of Linux via pushing aside Sysv init which had been the standard since the early 1980s when Ken Thompson and Denis Riche started working on the original AT&T System IV UNIX kernel which then became SystemV and somewhere in there companies like Sun Microsystems, SGI and other came along and forked the AT&T code and added in their proprietary code chunks further on down the family tree the AT&T codebase was again forked into 4.3BSD and then Linus Torvalds came along and forked the SysV kernel into what we today call the Linux kernel.

Linus did far more than “fork.” In his frozen dormitory room, he created the germ of “something entirely new and original,” and literally thousands of people have added to it since.

However: especially in this chosen corner of the computing world: ”nothing stays the same.” Nothing.

AFAIK Linus took the Minix kernel source, but simplified it to a monolithic kernel (modules are directly linked in, without using another layer). Then added some re-engineered Unix functionality. He never got a SysV license or source code.

He did not need a license, because your "understanding" is wrong. He did not work from the sources ANY pre-existing sources. Period. His implementation used the same definitions for certain signals and standards, but no code. This information is well documented in multiple places, you should look it up. Some of it makes interesting reading!

1 members found this post helpful.

No Linus didn't use the Minix kernel source for Linux, just a old falsehood by ESR.