Embedded KV stores are as comparable to SQLite as Cassandra is to Postgres (i.e. lacking joins or other data modelling essentials), but I agree that embedding a database directly inside your app is interesting and opens up a world of possibilities for efficiently blending N+1 code and queries. It gets even more interesting to take it a step further and embed a networked-replica of a database à la Datomic Peers.
I haven't used SQLite until a few weeks ago because I've always had an Oracle or Postgres running and it was an easy choice. Recently, I needed a self contained solution for a Python project, and I wanted the SQL support with minimum installation and system dependencies. I tried SQLite - I'm absolutely amazed how well it worked for the task.
I don't know how SQLite scales for large data sets, but for a small tasks it's fantastic, and can be trivially ported to something like Postgres.
It was probably meant as a learning exercise. And for me it is that implementation details of the KV/heap layer with some meaningful transactional properties seem more interesting than how to build SQL/whatever database on top of that, which in comparison looks like mostly obvious but tedious SMOP.
True, I was being somewhat pedantic. Whenever I see the word "db" I can't help but think "DBMS". It's frustrating that "DBMS" is such a mouthful that nobody bothers using it anymore, and that KV store projects choose to (ab)use the DB suffix.
> implementation details of the KV/heap layer with some meaningful transactional properties seem more interesting
There is undoubtedly a real beauty in LMDB's data structures, and it is impressive to see how Rockset have re-engineered RocksDB to become cloud-native …but my own feeling is that the SQL/whatever, distributed consistency and federation layers of a DBMS encompass some seriously hard and fascinating problems. What if the web could behave like a cohesive offline-first database for all of human knowledge after all?
Now I think about it, "interesting" has a pretty loose definition too!
I've only used Badger as an exploratory side project but I've found it enjoyable insofar as it feels much easier to get set up than something like RocksDB
