Travis told me it's alright now. :D
On Thu, Jul 31, 2014 at 11:58 AM, Ufuk Celebi <[email protected]> wrote: > I've changed this. > > Could you checkout the current master and then report back? :) > > On 31 Jul 2014, at 10:40, Stephan Ewen <[email protected]> wrote: > > > Or try "boolean hasRequestedClose = > Whitebox.<Boolean>getInternalState(queue, > > "hasRequestedClose");" > > > > > > On Thu, Jul 31, 2014 at 10:35 AM, Ufuk Celebi <[email protected]> > wrote: > > > >> I think the problem is with the way I use PowerMock in the test: > >> > >> boolean hasRequestedClose = Whitebox.getInternalState(queue, > >> "hasRequestedClose"); > >> > >> Could you try whether changing it to Boolean works? > >> > >> Best, > >> > >> Ufuk > >> > >> On 31 Jul 2014, at 09:45, Robert Metzger <[email protected]> wrote: > >> > >>> Hi, > >>> are you using OpenJDK or OracleJDK? There is a unfixed bug in Oracle > JDK > >> 6, > >>> but I think that leads to a null pointer exception in the compiler. > >>> Travis was able to build the current master with OpenJDK6 yesterday: > >>> https://travis-ci.org/apache/incubator-flink/jobs/31142009 > >>> > >>> Robert > >>> > >>> > >>> > >>> On Thu, Jul 31, 2014 at 9:13 AM, Aljoscha Krettek <[email protected] > > > >>> wrote: > >>> > >>>> Hi Guys, > >>>> I just noticed that "mvn verify" is failing on master: > >>>> > >>>> [ERROR] Failed to execute goal > >>>> org.apache.maven.plugins:maven-compiler-plugin:3.1:testCompile > >>>> (default-testCompile) on project flink-runtime: Compilation failure > >>>> > >>>> [ERROR] > >>>> > >>>> > >> > /home/travis/build/aljoscha/incubator-flink/flink-runtime/src/test/java/org/apache/flink/runtime/io/network/netty/OutboundConnectionQueueTest.java:[157,70] > >>>> type parameters of <T>T cannot be determined; no unique maximal > instance > >>>> exists for type variable T with upper bounds boolean,java.lang.Object > >>>> > >>>> We had this problem before, does anyone remember what the problem is > >> with > >>>> java6 here? Also, I'm not filing a bug but posting this here since it > >>>> should be handled quickly because a failing build looks bad. > >>>> > >>>> -- > >>>> > >>>> aljoscha > >>>> > >> > >> > >
