SE-Radio Episode 348 Riccardo Terrell on Concurrency

Immutability

Immutability plays a crucial role in enhancing the safety of concurrent programming by eliminating the risks associated with shared state corruption. Riccardo Terrell explains that immutability ensures that once a value is assigned, it remains unchanged, thus allowing multiple threads to access it without conflicts 1. This approach is increasingly embraced by modern programming languages like Go and Rust, which introduce immutable structures by default, contrasting with traditional languages such as C and Java 1.

Now they're more and more introducing this concept, the language, because they realize that the best way to solve this kind of problem is from the roof, removing completely what could be the source of the problem, which is a mutable state.

--- Riccardo Terrell

Terrell emphasizes that while adopting immutability may initially challenge developers accustomed to imperative programming, the long-term benefits in concurrency are significant 1.

   

Transparency

Referential transparency is a key concept in functional programming that facilitates concurrency by ensuring that functions are free of side effects. Riccardo Terrell highlights that this property allows functions to be executed in any order without affecting the outcome, making them ideal for parallel execution 2. This characteristic not only aids in structuring programs for concurrency but also enables caching of function outputs, as the results remain consistent for the same inputs 2.

Functional programming doesn't have side effects and therefore you could take set functions and run them concurrent because they will not interfere with each other anyway because there are no side effects, right?

--- Felienne Hermans

Terrell compares this to breaking down complex problems into smaller, independent tasks that can be solved concurrently, akin to the principles of functional programming 2.