[jobsub] Grep sample jar is misleading

The grep sample provided with Job Designer is not a good example of a MapReduce job design. In the example, the JAR file is set to /user/hue/jobsub/examples/hadoop-examples.jar. However, this JAR file is not used because the map and reduce classes that are used in the job are provided by the org.apache.hadoop.mapred.lib package. I have substituted a random JAR file in the JAR path field and the example still works.