On Mon, Jul 6, 2009 at 2:50 PM, tim robertson<[email protected]> wrote: > Thanks. I realised it didn't build out the box so I used 0.19.0 which > built ok, but I haven't run yet. > > So is this the recommended approach? > > $ svn co http://svn.apache.org/repos/asf/hadoop/hive/trunk hive > $ cd hive > $ ant -Dhadoop.version="0.19.1" package > > Thanks for the pointer to the issue - I might try and build with the > patch. (My motivation for 0.20.0 is to get the latest HBase on the > same HDFS) > > Thanks, > > Tim > > > > On Mon, Jul 6, 2009 at 6:07 PM, Edward Capriolo<[email protected]> wrote: >> On Mon, Jul 6, 2009 at 11:20 AM, tim robertson<[email protected]> >> wrote: >>> Hi All, >>> >>> So hive was a standalone project, then in 0.19.0 I saw it was in the >>> hadoop package but I never used it and now I see in >>> http://issues.apache.org/jira/browse/HADOOP-4628 it is a sub project >>> of Hadoop. >>> >>> Can I please ask if the instructions on >>> http://wiki.apache.org/hadoop/Hive/GettingStarted are likely to work >>> for Hadoop 0.20.0 ? >>> >>> >>> Cheers, >>> >>> Tim >>> >> >> Tim, >> >> There are a few upstream changes in Hadoop 0.20.0 that are breaking >> the Hive build. See this... >> https://issues.apache.org/jira/browse/HIVE-487 >> >> Hadoop Hive has broken off from Hadoop and your probably should not >> use the contrib version. >> >> Edward >> >
That approach is fine. You can check out the trunk or a release (3.0). The pro of 3.0 is with debugging you will have a point of reference for debugging help. Trunk will have the latest and greatest. (I am using a feature today from trunk :)
