Question

I currently program in Futures, and I’m rather curious about actors. I’d like to hear from an experienced voice:

• What are the advantages of actors over futures?
• When should I use one instead of other?

As far as I’ve read, actors hold state and futures doesn’t, is this the only difference? So if I have true immutability I shouldn’t care about actors?

Please enlighten me 🙂

Answer 1

One important difference is that actors typically have internal state, and therefore theoretically, they are not composable; see this and this blog post for having some issues elaborated. However, in practice, they usually provide a sweet spot between the imperative and the purely functional approach. So if possible, it is recommended to stick to programming with only futures, but if the message-passing model fits your problem domain better, feel free to use actors.