[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Systemd, systemd, and more systemd ...



Seems to be dominating a good part of the list traffic.

Personal history: I never had qualms about upgrading the old sysv stuff. If 
something did not work, it was some script which I could a least try to fix. 
The whole init sequence was RUNPARTS off various run-level directories 
containing appropriately named/numbered symlinks to scripts.

When this went over to a makefile-style precedence, so, some stuff would no 
longer work but the upgrade itself balked at this point. A decent decision was 
made to simply run those scripts at the end so thing proceeded. One of my own 
scripts stopped working. I had to tell what it must precede. Very difficult. 
Still serviceable using a text-editor or simple tools provided.

Systemd changed all that. So now, I am afraid to upgrade (running Sid) lest 
the system become unbootable with no ready recourse. So checking bug reports 
very carefully, waiting before venturing forth. And this hits udev and now 
cups as well.

Aside from that, systemd works great, is lightening-fast, had no problems with 
it! So why all the fuss. If posting on Sid would mean despite the expected 
bugs, the thing will work and not cripple the upgrader (maybe too ambitious 
for unstable maintenance), then no problem going ahead with this. And if a 
boot command "init=/lib/sysvinit/init" will definitely yield a fallback (have 
it in my lilo.conf but have not actually needed to tried it), then maybe this 
can be laid to rest.


Reply to: