sigh
Pretty much everyone has moved away from MySQL and Postgres to SQLite for embedded work and Berkeley DB stopped being usefully developed for it's original lightweight use years ago.
Let's see ... take a perfectly good old fashioned hash-keyed DB and turn it into some monolithic embedded SQL bloatware and then try to punt it into the position that SQLite already occupies - small, fast, truly public domain and still maintained by a single person's vision - and then try to tell the world how great you new release is. Hmm.