I have been reading up a bit on Erlang and its approach to writing highly distributed algorithms via lightweight processes and message passing.
I also recall the Prolog in Erlang style discussions a while back and the proposal suggested by @torbjorn.lager.
In practice can a highly performant system be envisioned where, Erlang is used at a low level via an FFI – i.e. tightly integrated – and Prolog for higher level symbolic computations.
I guess technically one would then run two VMs which seems like much overhead , for a combined system.
I am also wondering – what overhead does Erlang / OTP really have in its offering when compared to Prolog written in an Erlang style.
Any thoughts are much appreciated.