Good read: Right now, this is a distinction pretty much without a difference. If you choose an implementation of MapReduce — like pure Hadoop (say in the Cloud

Good read: Right now, this is a distinction pretty much without a difference. If you choose an implementation of MapReduce — like pure Hadoop (say in the Cloudera distribution) or Hadoop-Vertica or Aster Data’s SQL/MapReduce – you’re basically picking an entire technology stack. But those stacks are going to do a whole lot of changing and maturing in the near future – and as they do, it’s likely that projects will interact or even combine in all sorts of interesting ways.