On Thu, Jan 11, 2007 at 12:43:26AM +0200, Sami Haahtinen wrote: > Hmm.. That should be a bug. Atleast according to LSB documentation, > force-reload should be canonical to reload if it's available and restart > if it's not. > > http://www.freestandards.org/spec/booksets/LSB-Core-generic/LSB-Core-generic/iniscrptact.html > > I'd err on the side of safe, but in this case there is no safe side. =(
I verified the init.d script of autofs. And it is the case, reload exist and does a nice reload of the maps while force-reload is synonimous to restart, ie stop,start. So in that case, the proposed change would break it bad. Now, it might be a bug in autofs, I am not sure if the LSB core policy is supposed to be applied in debian. But what I am pretty sure of, is that there must be plenty of packages which break this policy. I think it is safer to make sure that the behavior is the one we want by coding it in the plugin rather than expecting the script to perform ok. And that does not prevent us from filling bug reports in the mean time. Now, when the usage line is given, the exit code should also be 1 right? I guess that this test should be pretty specific. We should not restart too many services twice by testing the exit code and the usage regexp. Did I miss something? jacques
signature.asc
Description: Digital signature