> Yeah, I'm guessing you're seeing systemd using /lib/init/rw is because > we ship a mount unit for that directory. Once all other packages have > stopped using it we can stop providing it. > > I see that the mount unit does not have a ConditionPathIsDirectory for > the unit, I'll fix that and then close this bug.
Adding ConditionPathIsDirectory is a good idea. It will basically make the unit a no-op as soon as /lib/init/rw goes away. Am 08.07.2011 18:55, schrieb Tollef Fog Heen: > ]] Roger Leigh > > | IIRC, I discussed with mbiebl whether or not we should add a versioned > | Breaks to initscripts on all packages currently using /lib/init/rw > | (once they are fixed to use /run) to ensure that all users of > | /lib/init/rw are no longer present, in order to allow the removal of > | /lib/init/rw for wheezy. Would it be appropriate for systemd to also > | use the same set of Breaks? > > I'd probably just depend on initscripts for now to not have to keep the > list in sync in two places. Unless we want to remove the lib-init-rw.mount unit file in time for wheezy, we can just rely on initscripts doing the cleanup for us and lib-init-rw.mount becoming a no-op (via ConditionPathIsDirectory) , and remove the file in wheezy + 1. That said, bumping the dependency on initscripts is not that big of a deal, so I guess we should just do that. Cheers, Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth?
signature.asc
Description: OpenPGP digital signature
