On Sat, Feb 11, 2012 at 1:46 AM, Karthik K <[email protected]> wrote: > > Curious, why we need to get both the jars (compiled against different > hadoop platform versions) in a single tar.gz ?
I did but do not think we need to have them in the same tar.gz. > How about doing the mvn compilation / assembly/ deploy lifecycle once for > each profile supported , to make sure they are indeed compiled against > different versions, packaged and are separate ? > > > And, people who want the appropriate versions can use the classifier to > pull the relevant libraries or download the package itself ? I believe I tried that but the maven repo didn't look "right". Beyond that, when I messed with classifiers the tar.gz did not contain the classifier name and thus would be overwrote? If you know the exact steps to make this proposal to work correctly, I'd be more than happy to go with this method. I just have no idea how to do this correctly. -- Apache MRUnit - Unit testing MapReduce - http://incubator.apache.org/mrunit/
