A propos, I actually find that the biggest win of FP is somewhat orthogonal to it, and that’s immutability. This leads to the ability to reason locally about code which is IMO completely invaluable even outside of parallelism scenarios. When you don’t know if a method might be mutating a parameter without actually reading all the way down it’s a real PITA 😅 (I have been doing a lot of refactoring of a large codebase recently and this has been a serious issue)
I don’t do pure FP, but do value the scala concept to write immutable by default unless there’s a good reason not to (usually efficiency, inner-loop stuff). While refactoring old code (most was originally java) I gradually convert vars to vals, but the old mutable code still works too, it’s multi-paradigm. What matters most for reasoning is readability, scala3 really helps with this.
A propos, I actually find that the biggest win of FP is somewhat orthogonal to it, and that’s immutability. This leads to the ability to reason locally about code which is IMO completely invaluable even outside of parallelism scenarios. When you don’t know if a method might be mutating a parameter without actually reading all the way down it’s a real PITA 😅 (I have been doing a lot of refactoring of a large codebase recently and this has been a serious issue)
I don’t do pure FP, but do value the scala concept to write immutable by default unless there’s a good reason not to (usually efficiency, inner-loop stuff). While refactoring old code (most was originally java) I gradually convert vars to vals, but the old mutable code still works too, it’s multi-paradigm. What matters most for reasoning is readability, scala3 really helps with this.