TL;DR: known example of a package correctly solving sysV+.service+.socket and having only .socket running on install?
- Detail - Hello everybody, I want to know if there is any best practise / example on the following case: A package that delivers: - sysV init (for old style compat/backport or just happens to be there) - .service supposed to replace the sysV - the .service should be installed and enabled, but not started in postinst - .socket installed and started on install supposed to be starting the service when needed I've hit a few cases like that now, and in most of them I see - dh_installinit for sysV trumps dh_installsystemd/dh_systemd_* - invoke.rc start is mapped to systemd service start - service ends up started which it shoudl not I have tried different combinations of --no-start and dh_*_override's, but never succeeded to something great so far. Different compat levels have great impact on this as well, unfortunately none resolved it for me. The few cases I know ended up dropping the sysV init to get out of the situation :-/ This is why I wanted to reach out if anybody knows better "prior art" to resolve this. -- Christian Ehrhardt Software Engineer, Ubuntu Server Canonical Ltd -- ubuntu-devel mailing list [email protected] Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
