Thread Pool is meant to be an easy to use implementation of a fairly naive thread pool.  For example, here's some code that will download 100 google search queries in parallel.

require 'open-uri'
require 'thread-pool'

number_of_files = 100
number_of_files.times do |i|
	ThreadPool.give_work("http://www.google.com/search?q=#{i}",i) do |url,index|
		data = open(url).read
		f = File.new("google-#{i}",'w')
		f.puts(data)
		f.close
	end
end

ThreadPool.start_up(number_of_files)

while ThreadPool.num_executing > 0 or ThreadPool.queue_size > 0
  sleep(1)
end

This approach of course has its limitations, and it's important to follow best practices with respect to share resources etc..., but it does make it very easy to distribute heavily I/O bound task across a number of non-blocking processes.