wow, too many uses to track down and too many commits to back out. Looks like Suneel and Stevo have been working on this? To support Hadoop 1.2.1 several things need to be backed out. I don’t have the time to work through them right now. I changed all the IsDirectory but isDir isn’t recognized and some other issues are coming up.
I’m going to have to move back before all those commits to get a client project going. On Mar 27, 2015, at 10:35 AM, Pat Ferrel <[email protected]> wrote: OK, if it’s just isDirectory I may do it and get a test on a 1.2.1 cluster to boot. On Mar 27, 2015, at 10:24 AM, Suneel Marthi <[email protected]> wrote: +1 with Java 7 will stick with Hadoop 1.2.1 support. Amen! On Fri, Mar 27, 2015 at 1:19 PM, Pat Ferrel <[email protected]> wrote: > ok, don’t see a major problem with forcing Java 7. > > So are we agreeing to support Hadoop 1.2.1? That would be a big thing to > drop even though I agree we must support 2.x and would come down in it’s > favor if forced to choose. > > > On Mar 27, 2015, at 10:16 AM, Suneel Marthi <[email protected]> > wrote: > > Its not about removing Guava, but trying to minimize Guava usage with the > upgrade to Java 7. Sure u can still use BiMaps but there are other things > like the guava Closeables that Stevo mentioned should be replaced with > equivalent Java 7 constructs. > > Switching isDir() to isDirectory() was my fault, will revert that back. > > On Fri, Mar 27, 2015 at 1:11 PM, Pat Ferrel <[email protected]> wrote: > >> BTW guava is in scala cooccurrence so not sure it’s going to be removed >> completely. It’s also used in Hadoop and Spark. Why are we trying to > remove >> it when it seems to be the defacto standard? Cooccurrence uses a unique > to >> Guava class (HashBiMap) so to remove it is not trivial. I’ve seen scala >> versions but why fragment into implementing out own generic collections. > Is >> there a HashBiMap in Java 7, I haven’t heard. >> >> Anyway I don’t think forcing an upgrade to Java 7 is all that big a deal. >> Go ahead and remove guava is you want but be aware that one part is still >> being used. >> >> I’m more concerned with dropping support for hadoop 1.2.1 without >> compelling reasons. >> >> On Mar 27, 2015, at 9:57 AM, Suneel Marthi <[email protected]> >> wrote: >> >> Agree with Stevo. there's another round of code cleanup we need to go > thru >> to eliminate guava collection calls and replace with straight Java 7 > calls. >> We agreed as part of the last hangout that Java 7 would be the minimal > Java >> that we support going forward. >> >> On Fri, Mar 27, 2015 at 12:51 PM, Stevo Slavić <[email protected]> > wrote: >> >>> Question is what is minimal Java that we should require Mahout users to >>> use. It was raised to 1.7 via MAHOUT-1652 >>> <https://issues.apache.org/jira/browse/MAHOUT-1652>. It enables us to >> use >>> Java 1.7 APIs, finally. E.g. after level was raised I was recently able >> to >>> eliminate some usage of Guava for closing Closable's with using >>> try-with-resources Java 7 construct. We should minimize dependencies, >> there >>> are IMO too many. Depending more and more on standard libraries of Java >> and >>> Scala helps in that direction. Hopefully we do not wait much longer >> before >>> the level is raised even further to 1.8, so we have even less 3rd party >>> dependencies. >>> >>> On Fri, Mar 27, 2015 at 5:31 PM, Pat Ferrel <[email protected]> >> wrote: >>> >>>> It should, Hadoop supports it long term and lots of people stuck there >>>> with projects that haven’t been upgraded (Mahout comes to mind). >>>> >>>> On Mar 27, 2015, at 9:26 AM, Stevo Slavić <[email protected]> wrote: >>>> >>>> Have to check but I doubt that build supports hadoop 1.x any more. >>>> >>>> On Fri, Mar 27, 2015 at 5:15 PM, Suneel Marthi < > [email protected] >>> >>>> wrote: >>>> >>>>> This is the Java version, gotta use Java 7 >>>>> >>>>> On Fri, Mar 27, 2015 at 12:08 PM, Pat Ferrel <[email protected]> >>>>> wrote: >>>>> >>>>>> Latest source for Spark 1.1.0 and Hadoop 1.2.1 >>>>>> >>>>>> Build complains about the move to >>>>>> <maven.compiler.target>1.7</maven.compiler.target> I think this was >>>> upped >>>>>> from 1.6 but not sure if that’s what the error is about. >>>>>> >>>>>> I’m on Java 6 no this machine if that matters. >>>>>> >>>>>> Actual error: >>>>>> >>>>>> [INFO] Mahout Build Tools ................................ SUCCESS >>>>> [3.512s] >>>>>> [INFO] Apache Mahout ..................................... SUCCESS >>>>> [0.603s] >>>>>> [INFO] Mahout Math ....................................... FAILURE >>>>> [6.453s] >>>>>> [INFO] Mahout MapReduce Legacy ........................... SKIPPED >>>>>> [INFO] Mahout Integration ................................ SKIPPED >>>>>> [INFO] Mahout Examples ................................... SKIPPED >>>>>> [INFO] Mahout Release Package ............................ SKIPPED >>>>>> [INFO] Mahout Math Scala bindings ........................ SKIPPED >>>>>> [INFO] Mahout Spark bindings ............................. SKIPPED >>>>>> [INFO] Mahout Spark bindings shell ....................... SKIPPED >>>>>> [INFO] Mahout H2O backend ................................ SKIPPED >>>>>> [INFO] >>>>>> >>> ------------------------------------------------------------------------ >>>>>> [INFO] BUILD FAILURE >>>>>> [INFO] >>>>>> >>> ------------------------------------------------------------------------ >>>>>> [INFO] Total time: 11.609s >>>>>> [INFO] Finished at: Fri Mar 27 08:55:35 PDT 2015 >>>>>> [INFO] Final Memory: 24M/310M >>>>>> [INFO] >>>>>> >>> ------------------------------------------------------------------------ >>>>>> [ERROR] Failed to execute goal >>>>>> org.apache.maven.plugins:maven-compiler-plugin:3.2:compile >>>>>> (default-compile) on project mahout-math: Fatal error compiling: >>> invalid >>>>>> target release: 1.7 -> [Help 1] >>>>>> [ERROR] >>>>>> [ERROR] To see the full stack trace of the errors, re-run Maven with >>> the >>>>>> -e switch. >>>>>> [ERROR] Re-run Maven using the -X switch to enable full debug > logging. >>>>>> [ERROR] >>>>>> >>>>>> >>>>> >>>> >>>> >>> >> >> > >
