This looks amazing! This appears to perfectly fit the huge gulf between pl/pgsql and C in the Postgres stack.
Pl/pgsql, while not exactly elegant, is well suited to bridging that gap between imperative and set-oriented business logic within the database.
On the other hand, pl/pgsql is far from optimal for defining custom types and implementing operators for those types. For that, we've needed C.
But now the possibility opens up to create efficient custom types, operators, and other functions that can run at machine speed without depending on what a vendor decides to bundle or the mostly unvetted quality of some 3rd party's C code extension.
Pl/pgsql, while not exactly elegant, is well suited to bridging that gap between imperative and set-oriented business logic within the database.
On the other hand, pl/pgsql is far from optimal for defining custom types and implementing operators for those types. For that, we've needed C.
But now the possibility opens up to create efficient custom types, operators, and other functions that can run at machine speed without depending on what a vendor decides to bundle or the mostly unvetted quality of some 3rd party's C code extension.
Potentially huge!