GPars 0.11 adds dataflow concurrency to its existing intuitive ways to handle tasks, actors and messages in Groovy. In addition, this new release improves performance for actors and provides a Java API.

GPars, formerly GParallelizer, is also known as Groovy Parallel Systems. This open source project has been evolving from a simple Scala-like actors implementation to a very complete parallel programming DSL (short for Domain-Specific Language) that supports the most popular concurrency concepts. Its newest release, GPars 0.11 provides support for the following abstractions:

• Actors
• Tasks
• Dataflow tasks, variables, streams and operators
• Asynchronous programming
• Agents
• Message-passing
• Concurrent collections

One of the great problems of the previous releases was the poor performance for actors. GPars 0.11 has achieved important speedups in the actors messaging core. Now, it is more attractive to experiment with actors with GPars. This new release also improved performance for agents. If you have existing code designed for previous GPars versions, you should consider the upgrade because you will notice very important speedups without changes in your code.

Dataflow concurrency is becoming popular because it makes it simple to avoid race conditions and other concurrency nightmares. In addition, dataflow concurrency usually allows you to create code that is easy to read and understand. If you are interested in dataflow concurrency with Groovy, you can read the details for the dataflow constructs implemented in GPars 0.11 in this simple Dataflow Programming Guide. The guide includes easy-to-understand sample code that uses many of the new dataflow constructs.