On 17/11/2014 15:14, Paul H wrote: > My plan is that if I find myself depending on packages which no > longer work under sysvinit-core, I'll rebuild those packages for > myself as needed (and share the results, if that's helpful): I have > to do this already now anyway, for example to use build options which > make more sense for a given target, or to make my own -nox (headless) > versions of stuff that would otherwise drag in 200MiB+ of GTK/Qt > dependencies.
I like this approach, and I think it is a good way to go. I guess the only downside is that you end with a bunch of debuild recipes that are functionally equivalent to a folder of full of .bb files :) I think that in practice the amount of packages that will actually depend on systemd being PID1 will stay relatively small. I have no illogical problem with pulling in libsystemd and its friends. >>> it seems that realtime and systemd is problematic. Has anyone >>> tried the workarounds mentioned with a PREEMPT_RT kernel? How did >>> it go? > > It is my intention to make the |ControlGroup=cpu:/| approach work; I > have been maintaining a systemd rootfs for one of my targets but > haven't quite found the time to investigate this fully. Let us know what you find. Thanks, Robert -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/546a824f.5000...@gmail.com
