Interestingly, the version in Maven is 2.9.1, but SBT outputs file to the 2.9.1.final
directory inside target.

A couple of small changes in SparkIMain were also required.

All tests pass and ./spark-shell launches successfully.

by specifying per-executor memory.

The repl module requires more complex work.

Replacing the native lzf compression code with the ning open-source compress-lzf library. (Apache 2.0 liscense)

directory by setting SPARK_HOME and locating the executor relative to
that. Entries on SPARK_CLASSPATH and SPARK_LIBRARY_PATH are also passed
along to worker nodes.

end. The same problem is seen also in the master branch (in the
ChainedStreaming implementation)

Changed printlns to log statements and fixed a bug in run that was causing it to fail on a Mesos cluster

that unit tests can run successfully.

in general through a conf/spark-env.sh script.

to select source instead of leecher count.

- Receivers now ask for a range instead of expecting the whole variable. But,
  they are still asking for the whole range from a single source.
- Next step: make receivers ask for different parts from different sources.
  Also, make sure that Master sends back a list of sources instead of a single
  one.