On Thu, Jul 22, 2010 at 12:36:34PM +1000, Dave Airlie wrote: > On Thu, 2010-07-22 at 04:25 +0200, Lennart Poettering wrote: > > i.e. the "enable"/"disable" commands makes some changes for the next > > time they are looked at, and then adding --realize on top makes those > > changes take effect immediately, i.e. so that the unit is start/stopped > > according to those changes. We actually used "--start=" first (which > > however is very confusing when you'd write "disable --start" to disable > > something and then have it stop...) We then considered "--now", because > > it is not a verb. But eventually we stuck with --realize. It's not > > great, yes. But we couldnt think of anything better. Happy to take > > suggestions. But no, --take-effect-immediately is not really an option. > > Why have two verbs in a command structure? isn't enable or disable the > order, --now seems like it would be correct, the thing with English is > its flexible about these sort of things. > > Also you have --realize=reload and --realize=minimal, again > non-representative, minimal means reload if running, whereas reload > means start if not running otherwise reload? I would expect reload means > reload if running, otherwise do nothing, the other option would probably > be better specified as reload,start. I notice that the man page takes great pains to distinguish enable/disable and start/stop (activate/deactivate). This makes sense -- the current commands separate into "chkconfig on/off" and "service start/stop". So why are the concepts being blurred here? There's already a systemctl start/stop/... (as mentioned in the systemd-install description). Maybe what you want is an --if-enabled option to systemctl start/restart/...? -- devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/devel
