Previously, 30 seconds after the last notification disappears the daemon exited. This timeout was hard-coded.
Issue
I can only assume on MATE that some other process is claiming the DBus name and spinning up the daemon whenever a notification comes through. Whatever that process is, it doesn't work when I run the the daemon on Sway without other MATE components running. I have to run it in a service that constantly restarts it, which is annoying.
Proposal
This PR keeps the 30 seconds as the default and adds a command line option that allows setting it to something else (notably -1 which will disable the timeout).
I now understand that DBus services are a thing and that's how mate-notification-daemon is expected to be run. With that information, this PR no longer makes sense.
Status quo
Previously, 30 seconds after the last notification disappears the daemon exited. This timeout was hard-coded.
Issue
I can only assume on MATE that some other process is claiming the DBus name and spinning up the daemon whenever a notification comes through. Whatever that process is, it doesn't work when I run the the daemon on Sway without other MATE components running. I have to run it in a service that constantly restarts it, which is annoying.
Proposal
This PR keeps the 30 seconds as the default and adds a command line option that allows setting it to something else (notably -1 which will disable the timeout).