Agent: Like an atom such that it will represent some state of an identity at some point in time, but can be asynchronous. For example you can keep sending work to an agent and carry on with your program.
Unlike the way Atoms synchronously accept changes to their identities through the compare-and-set mechanism, agents allow you to asynchronously send off work and that will block each other until their work is complete.
Interacting with an agent is very similar to using atoms. If we wanted to represent a checking account as an agent:
(def checking-account (agent {:balance 100 :holder "my-name"}))And changing the account holder:
(send! checking-account (comp #(assoc % :holder "your-name")))We can make slight modifications to our previous example in lib/agents.clj. Simply follow the directions provided to you in the comments. You'll notice that there are not many changes to be made in implementation, but will notice a subtle difference in behavior of your program.
TIP: skip ahead to an implementation of using delays with
git checkout agents-solved
How would you explain the differences between an Atom and an Agent? How are they similar? Can you give any examples of when you would want to use one over the other?
Continue on to Refs.