Improving the efficiency of tuple spases

Linda provides high-level linguistic abstractions for concurrent programming with operations for synchronization and exchange of values between different programs that share information by accessing common repositories named tuple spaces. Despite their expressive power and their simplicity, there are several challenges in implementing tuple space systems, which prevent the Linda model to be widespread. The goal of this work is to provide an efficient implementation of the Linda coordination model. As a starting point, we take KLAIM and its Java implementation and we improve on it after evaluating the performances of selected implementations of tuple spaces and discussing their different implementation choices concerned with data structures and querying techniques. Our KLAIM implementation is also extended with abstractions for data replication based on automatic data placement strategies.