Changed numbers all over, so that the progression is more natural

30-double-inverse-multiplexer/README.md

@@ -0,0 +1,22 @@
+Style #13
+==============================
+
+Very similar to style #12, but with an additional twist
+
+Constraints:
+
+- Input data is divided in chunks, similar to what an inverse multiplexer does to input signals
+
+- A map function applies a given worker function to each chunk of data, potentially in parallel
+
+- The results of the many worker functions are reshuffled in a way
+  that allows for the reduce step to be also parallelized
+
+- The reshuffled chunks of data are given as input to a second map
+  function that takes a reducible function as input
+
+Possible names:
+
+- Double inverse multiplexer 
+- Map-reduce 
+- Hadoop style