The expectation around CR releases is that they are the Final release. It just so happened that CR1 had a major bug that I felt warranted another CR. After the first CR release I general point the version to the next point release. IMO pushing new 5.0.0 SNAPSHOTs after a 5.0.0 Final or CR is inaccurate. So yes, 5.0.1-SNAPSHOT is the expected current master version. Its only really an issue in cases like this where CR1 has such a major bug. Scott just uses the built jars directly iiuc rather than relying on Maven/Gradle/etc resolution, so he would just do a `gradle jar` and grab the jars. I do not have this problem running releases nor do the CI jobs have this problem. These both do a publish. So I find this weird. It would seem to be a task dependency problem. Try building hibernate-osgi first, and then running the publish. On Sun, Jul 5, 2015 at 2:45 PM Sanne Grinovero <sanne(a)hibernate.org&gt; wrote: > Hi all, > I'm trying to build Hibernate ORM's latest to try debugging some of > the latest changes for Wildfly and Search integrations. > > The README mentions I should be able to > > ./gradlew clean publishToMavenLocal > > but that fails on the hibernate-osgi module. I then tried "publish" to > upload a snapshot, but that prouces the same error message: > > > :hibernate-osgi:publishMavenJavaPublicationToJboss-snapshots-repositoryRepository > FAILED > > FAILURE: Build failed with an exception. > > * What went wrong: > Execution failed for task > > ':hibernate-osgi:publishMavenJavaPublicationToJboss-snapshots-repositoryRepository'. > > Failed to publish publication 'mavenJava' to repository > 'jboss-snapshots-repository' > > Invalid publication 'mavenJava': artifact file does not exist: > > '/home/sanne/workspaces/hibernate/hibernate-core-parent/hibernate-osgi/target/karafFeatures/hibernate-osgi-5.0.1-SNAPSHOT-karaf.xml' > > On the version, I was expecting it to build some "5.0.0-SHAPSHOT" but > it's actually building a "5.0.1-SNAPSHOT", is that expected? > > I was hoping I could ignore the osgi module and take what it built to > experiment with Scott's branch, but while on HHH-9887 he mentions > being down to two failures, I'm having still dozens of failures so I > suspect I'll need some advice on the build. > > Scott, could you share which exact commit id did you use to test > Hibernate ORM on your branch "hibernate5_july2", and how I can build > it like yours ? > > Thanks, > Sanne > _______________________________________________ > hibernate-dev mailing list > hibernate-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/hibernate-dev >

> I was hoping I could ignore the osgi module and take what it built to > experiment with Scott's branch, but while on HHH-9887 he mentions > being down to two failures, I'm having still dozens of failures so I > suspect I'll need some advice on the build. > > Scott, could you share which exact commit id did you use to test > Hibernate ORM on your branch "hibernate5_july2", and how I can build > it like yours ? To build ORM, I use: ./gradlew clean install generateKarafFeatures publishToMavenLocal I last built against ORM master commit id 9b5bb9751cdd15eaf1049d2b51f6f4a61b878ed8 Then I built WildFly via: cd wildfly /build.sh clean install -DskipTests=true To recreate the WildFly test failures: cd wildfly/testsuite/integration/basic mvn clean install -Dtest=*JPA* > > Thanks, > Sanne > _______________________________________________ > hibernate-dev mailing list > hibernate-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/hibernate-dev > _______________________________________________ hibernate-dev mailing list hibernate-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/hibernate-dev