worker_pool ... extended

[Dmitrii Dimandt]

I've somewhat extended RabbitMQ's excellent implementation of worker pools (see full description at http://www.lshift.net/blog/2010/03/29/on-the-limits-of-concurrency-worker-pools-in-erlang)

Limitations of RabbitMQ's worker pool are as follows:

	• only one can be created within the running application
	• you have to supply the pool with the function you want to run each time you invoke it

My fork allows you to:

	• Run several worker pools (each independently named, if needed)
	• Make a worker pool execute a specified function every time it's invoked
		• If such a function requires initialization of certain parameters or somesuch, you can supply an init function that will be called once the worker pool has been started

You can get it at http://github.com/dmitriid/worker_pool

There are examples in the README, and the code should be working (despite the fact that I pulled it from a different project)