+1 Do I understand correctly, with this proposal there will be no common only committers? That is, there will be no committers to common who are not either HDFS or MR committers.
--Konstantin On 8/16/2010 2:38 PM, Owen O'Malley wrote:
Since the project split, it is mostly required that any MapReduce or HDFS committers have access to change Common. We haven't had any cases where we wanted to nominate any one for just Common. Toward the goal of simplifying the structure, I'd propose that we define the Common committers as precisely the union of the HDFS committers and MapReduce committers. Clearly, I'm +1. -- Owen
