Hi guys,
Two quick questions: 1. Are there any performance gains from hadoop streaming or pipes ? As far as I read, it is to ease testing using your favorite language. Which I think implies that everything is translated to bytecode eventually and executed.