Important anti-systemd essay ... please explain

nss000

Distinguished
Apr 18, 2008
673
0
19,010
Gents:

Trying to thoughtfully choose/reject the upgrade to non-systemd U_14.04 I came across this essay ....

http://ewontfix.com/14/

It's seems important, thought more than a little beyond my real understanding. Will one of the able Linux system THW posters identify the telling points and state them plainly?
 
the top comment summarizes it well.

This post remembers me of microkernel discussions. The message is: putting too many things in critical PID 1 is dangerous!
"Unfortunately, it also gets the other properties, including bringing down the whole system when it crashes."
To avoid this, he tells us that we must turn PID 1 into mostly a execve() call, and move the complexity to other layers.
But if you move all the PID 1 functionality to PID 2, then PID 2 becomes your critical point of contention. You haven't got rid of complexity, you just have moved the complexity around.
The "criticality" of the init system is not in PID 1, it's on the init system. Obsessing with keeping PID 1 void of all functionality without putting attention to anything else is not very clever.
 

nss000

Distinguished
Apr 18, 2008
673
0
19,010
BigS:

And then I found **this**:

http://www.bioscentral.com/misc/biosbasics.htm

Which makes it very clear that all serious engineering process has already been performed by the time an/any/the OS is perked into awareness. Might-as-well have hardwire chip-BIOS do it all ... each component looking after its own business, each application software self-contained and a FPGA attaching all hardware pieces. Why indeed have an OS and an eggo-fluffed SYSTEMD ?