Title :
Stage: Python with Actors
Author :
Ayres, John ; Eisenbach, Susan
Author_Institution :
Dept. of Comput., Imperial Coll. London, London
Abstract :
Programmers hoping to exploit multi-core processors must split their applications into threads suitable for independent, concurrent execution. The lock-based concurrency of many existing languages is clumsy and error prone - a barrier to writing fast and correct concurrent code. The Actor model exudes concurrency - each entity in the model (an Actor) executes concurrently. Interaction is restricted to message passing which prevents many of the errors associated with shared mutable state and locking, the common alternative. By favouring message passing over method calling the Actor model makes distribution straightforward. Early Actor-based languages enjoyed only moderate success, probably because they were before their time. More recent Actor languages have enjoyed greater success, the most successful being ERLANG, but the language is functional; a paradigm unfamiliar to many programmers. There is a need for a language that presents a familiar and fast encoding of the Actor model. In this paper we present STAGE, our mobile Actor language based on PYTHON.
Keywords :
concurrency control; message passing; mobile computing; object-oriented languages; ERLANG; Python programming language; STAGE mobile actor language; actor model; actor-based language; independent concurrent execution; lock-based concurrency; message passing; multicore processor; Application software; Concurrent computing; Educational institutions; Error correction codes; Message passing; Mobile computing; Multicore processing; Pervasive computing; Programming profession; Yarn;
Conference_Titel :
Multicore Software Engineering, 2009. IWMSE '09. ICSE Workshop on
Conference_Location :
Vancouver, BC
Print_ISBN :
978-1-4244-3718-4
DOI :
10.1109/IWMSE.2009.5071380