Authors: Paolo Baldini, Ylenia Battistini
The project, developed for the Concurrent and Distributed Programming course at University of Bologna, is about the development of a multiplayer version of the Mastermind game.
The system is based on the actor concept and use for his implementation the Akka framework.
Being a study-project, it uses some functionality not required for the development of the system (e.g., both classic and typed actors, receptionist, etc.), but added for a better comprehension of the framework/paradigm.
The game is a variant of the famous game Mastermind, where a player should guess a secret code created by a code-maker.
This version of the game is however a multiplayer one, where a player should guess all the adversarys' codes. The whole system is administrated by an Arbiter, who take some decisions such as the turn assignment, the ban of a player (if he think to have won, when instead he doesn't), etc.
Due to a possible high number of player and a length of codes that can vary, a solution like the one used in the Five-guess algorithm (who use minimax technique) was not feasible in terms of memory allocation. Due to that a new algorithm was implemented. This one does not focus on resolve the code in the minimun number of moves, but on the use of a lazy calculation for the next code to try. In fact, there is no need that a code must be discovered in a determined number of turn, but there is the necessity to use less memory as possible.
In this paragraph, we don't desire to accurately describe how messages are passed, in which order, or the fault techniques adopted. We are only interested in show a macro-structure of the message-passing between the actors and the GUIs frames. That will be done through use of the following image.
