Fleshing out the README

README

@@ -1 +1,29 @@
-This is a simple library demonstrating the analysis of the CommonCrawl dataset.
+This is a simple library demonstrating the analysis of the CommonCrawl dataset
+through implementing the canonical Hadoop Hello World program, a simple word
+counter.
+
+To build
+--------
+
+You'll need to have Apache Ant (http://ant.apache.org/manual/install.html)
+installed, and once you do, just run a:
+
+# ant dist
+
+This step will compile the libraries and Hadoop code into an Elastic MapReduce-
+friendly JAR at dist/lib/HelloWorld.jar, suitable for use as a custom JAR-based
+Elastic MapReduce workflow.
+
+To run locally
+--------------
+
+You'll need to be running Hadoop, and if you don't have it installed, Cloudera
+provides a useful set of OS-specific Hadoop packages which will make it easy.
+Check out their site:
+
+https://ccp.cloudera.com/display/SUPPORT/Downloads
+
+Once you've got Hadoop installed, you can use the 'hadoop jar' task to execute
+the tutorial code.  Here's the pattern:
+
+hadoop jar <checkout location>/dist/lib/HelloWorld.jar org.commoncrawl.tutorial.HelloWorld <Amazon AWS access key ID> <Amazon AWS secret access key> <CommonCrawl crawl files to use as input> <HDFS output location>