@zwol Yeah, this does feel like a very systemd-style approach, with all the controversy that implies 😅
I guess fundamentally though: the plain-text databases in /etc haven't been sufficient for decades, people need to dynamically generate answers to those queries, and expecting every process to run its own independent copy of whatever code the sysadmin has configured is the opposite of "do one thing well".
So really, what's more "Unix philosophy" than splitting these lookups out to separate daemons? 🤔