Re: existing alternatives
Svante Signell:
The more important that Debian does not drop support for sysvinit
> then, until alternatives have stabilized :) (and systemd/uselessd is
> deferred to PID 2). PID 1 should be as small as possible, see a
> proposed implementation in: http://ewontfix.com/14/
That proposed implementation is misleading and clearly not used in
practice. Moreover, alternatives exist, and have already existed for
years. Instead of looking at that hypothetical, go and look at some
actual implementations of process #1 programs to see what working
programs, including several minimalist ones, do. See:
* Gerrit Pape's runit (http://smarden.org/runit/runit.8.html)
(https://packages.debian.org/jessie/runit)
* Felix von Leitner's minit (http://www.fefe.de/minit/)
(https://packages.debian.org/jessie/minit)
* John Fremlin's jinit (http://john.fremlin.de/programs/linux/jinit/)
* Nico Schottelius' cinit (http://www.nico.schottelius.org/software/cinit/)
* Joachim Nilsson's finit (http://troglobit.com/finit.html)
* Init NG (http://initng.org/trac)
* the system manager in the nosh package
(http://homepage.ntlworld.com./jonathan.deboynepollard/Softwares/nosh.html)
Reply to: