Published on 26 May 2014.
I continued reading about Erlang this weekend. I read about the sequential part which is functional. It feels like Erlang programs are structured as small functional processes that communicate with message passing. I have to read more about the concurrent part of Erlang to know for sure though.
This feels similar to an approach I read about (I forgot where) when writing classes in object oriented programming: have all private methods be free of side effects. So each class might have a small amount of state that is only modified at the top level, but all the private “helper” methods would be purely functional.
Another thing it feels similar to is functional core imperative shell by Gary Bernhardt.
Because Erlang is a functional language, it is probably easier to enforce the functional part of such a design. It feels like Erlang is a good fit for writing programs the way some people like to write object oriented programs. Which is a bit strange.
Site proudly generated by Hakyll.