I am a Postgres advocate (as opposed to MySQL/MariaDB for contemporary use), but this article about using Postgres as a job queue (with near-realtime job claiming, no less!) really highlights why I hold this philosophy. MySQL and its derivatives might work well as a raw data store if you're willing to put in the effort to guard against accidental data corruption or tweak the compatibility settings just right, but Postgres, even though sometimes a bit feature-creeped at times,¹ provides for any data storage use case, and most are accounted for out of the box.

¹ Indeed I don't think that pub/sub would be an integral part of an RDBMS and it's a bit inexplicable that Postgres has it, but it nonetheless can be convenient in use cases such as presented within that article.