[GitHub] incubator-spark pull request: Patch for SPARK-942

[ash211]

Github user ash211 commented on the pull request:

    https://github.com/apache/incubator-spark/pull/180#issuecomment-34660329
  
    I often use Spark for ETL and the ability to stream directly onto (HDFS) 
disk would be valuable for me.  No need to buffer in memory if we don't have to.
    
    Admins, is there anything else you want to see from this PR before merging?