Fully agree. We should create these features during the Apache Aries
releases. This keeps our karaf features small and
allows people to use new Aries versions without switching the karaf version.
Christian
On 05.02.2016 12:43, Jean-Baptiste Onofré wrote:
Hi Jochen
thanks for the update. I think it makes sense to provide multiple
feature version (stable) and document which one to use depending of
the use case (annotation, blueprint, etc).
WDYT ?
Regards
JB
On 02/05/2016 12:19 PM, jochenw wrote:
An update for Karaf 4.0.4:
Using JPA 1.0.4 with org.apache.aries.jpa.container leads to a
problem with
finding the datasources:
org.apache.aries.jpa.container - 1.0.3 | No JNDI services for
jdbc/myXADataSource are available from persistence bundle
mypersistencebundle/0.1.2.SNAPSHOT. The persistence units [] will now be
unregistered.
Switching back to 1.0.2 in the enterprise feature XML solves that.
Using eclipselink 2.6.1 with org.apache.aries.jpa.eclipselink.adapter
2.3.0
leads to the the same problem as with the previous version:
java.lang.NoClassDefFoundError:
org/eclipse/persistence/internal/jpa/rs/metadata/model/Link
Switching back to 2.1.0 in the enterprise feature XML solves that.
Does it make sense to file a Jira for this, or is the JPA 1.x branch
sort of
deprecated, and we should switch to 2.x anyways, switching to the
annotation
way for defining persistence units?
Best Regards,
Jochen
--
View this message in context:
http://karaf.922171.n3.nabble.com/Karaf-4-0-2-Problem-Eclipselink-2-6-0-with-JPA-2-2-0-tp4043068p4045198.html
Sent from the Karaf - User mailing list archive at Nabble.com.
--
Christian Schneider
http://www.liquid-reality.de
Open Source Architect
http://www.talend.com
