For reference, in 1.14.1-0ubuntu8.1 you have this:
# Automatically added by dh_installinit
if [ -e "/etc/init/apport.conf" ]; then
# start fails if already running
start apport || :
fi
# End automatically added section
while in 1.20.1-0ubuntu3 you have this instead:
# Automatically added by dh_installinit
if [ -e "/etc/init/apport.conf" ]; then
invoke-rc.d apport start || exit $?
fi
# End automatically added section
I'm not sure, but that looks like a regression to me, moving away from upstart's "start" to "invoke-rc.d", so it may have come from old code. The old code hypothesis is kind of confirmed by the bad idea of exiting with a non-zero code in a postinstall script for no good reason when the 10.10 package just didn't care if the script failed.
For reference, in 1.14.1-0ubuntu8.1 you have this:
# Automatically added by dh_installinit apport. conf" ]; then
if [ -e "/etc/init/
# start fails if already running
start apport || :
fi
# End automatically added section
while in 1.20.1-0ubuntu3 you have this instead:
# Automatically added by dh_installinit apport. conf" ]; then
if [ -e "/etc/init/
invoke-rc.d apport start || exit $?
fi
# End automatically added section
I'm not sure, but that looks like a regression to me, moving away from upstart's "start" to "invoke-rc.d", so it may have come from old code. The old code hypothesis is kind of confirmed by the bad idea of exiting with a non-zero code in a postinstall script for no good reason when the 10.10 package just didn't care if the script failed.