[Hibernate-JIRA] Created: (ANN-712) Maven pom.xml should (in time) depend on javax.persistence:persistence-api instead of org.hibernate:ejb3-persistence
4:34 a.m.
Maven pom.xml should (in time) depend on javax.persistence:persistence-api instead of
org.hibernate:ejb3-persistence
--------------------------------------------------------------------------------------------------------------------
Key: ANN-712
URL: http://opensource.atlassian.com/projects/hibernate/browse/ANN-712
Project: Hibernate Annotations
Issue Type: Task
Affects Versions: 3.3.1.GA
Reporter: Geoffrey De Smet
Priority: Minor
Now it's possible to get the same classes (for example javax.persistence.Entity) twice
in the classpath:
- one from persistence-api-1.0.jar
- and another from ejb3-persistence-1.0.1.GA.jar.
Mix that with classloaders and you get bugs like A extends B but an A is not instance of B
(because B.class is loaded twice from different jars). Otherwise you can get this too, but
if the groupId:artifactId match, maven deals with it.
In theory, replace:
<dependency>
<groupId>org.hibernate</groupId>
<artifactId>ejb3-persistence</artifactId>
<version>1.0.1.GA</version>
</dependency>
with:
<dependency>
<groupId>javax.persistence</groupId>
<artifactId>persistence-api</artifactId>
<version>1.0</version> <!-- FIXME -->
</dependency>
It looks like org.hibernate:ejb3-persistence is just a copy of
javax.persistence:persistence-api, but with a hot-fix applied to it:
probably because javax.persistence:persistence-api:1.0.1 isn't deployed to the central
repo yet.
So the JCP might need to be poked to put javax.persistence:persistence-api:1.0.1 on the
central repo :) Which is probably not an easy thing to do...
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
10:10 a.m.
New subject: [Hibernate-JIRA] Resolved: (ANN-712) Maven pom.xml should (in time) depend on javax.persistence:persistence-api instead of org.hibernate:ejb3-persistence
[
http://opensource.atlassian.com/projects/hibernate/browse/ANN-712?page=co...
]
Emmanuel Bernard resolved ANN-712.
----------------------------------
Resolution: Rejected
The story is more complex than that. ejb3-persistence is not a fork of javax.persistence,
both evolved independently and are quite similar because the spec is the same :)
Sometimes, someone update a javax.* jar in the repo. But the TCK and the APIs evolve over
time to cope with TCK or Spec bugs.
I'd rather keep ejb3-persistence as a dependency. This is the version I use to compile
and the version passing the TCK for Hibernate and JBoss AS.
you should not have the javax.persistence dependency pulled if you use the pom I pushed,
AFAIK
Maven pom.xml should (in time) depend on
javax.persistence:persistence-api instead of org.hibernate:ejb3-persistence
--------------------------------------------------------------------------------------------------------------------
Key: ANN-712
URL: http://opensource.atlassian.com/projects/hibernate/browse/ANN-712
Project: Hibernate Annotations
Issue Type: Task
Affects Versions: 3.3.1.GA
Reporter: Geoffrey De Smet
Priority: Minor
Now it's possible to get the same classes (for example javax.persistence.Entity)
twice in the classpath:
- one from persistence-api-1.0.jar
- and another from ejb3-persistence-1.0.1.GA.jar.
Mix that with classloaders and you get bugs like A extends B but an A is not instance of
B (because B.class is loaded twice from different jars). Otherwise you can get this too,
but if the groupId:artifactId match, maven deals with it.
In theory, replace:
<dependency>
<groupId>org.hibernate</groupId>
<artifactId>ejb3-persistence</artifactId>
<version>1.0.1.GA</version>
</dependency>
with:
<dependency>
<groupId>javax.persistence</groupId>
<artifactId>persistence-api</artifactId>
<version>1.0</version> <!-- FIXME -->
</dependency>
It looks like org.hibernate:ejb3-persistence is just a copy of
javax.persistence:persistence-api, but with a hot-fix applied to it:
probably because javax.persistence:persistence-api:1.0.1 isn't deployed to the
central repo yet.
So the JCP might need to be poked to put javax.persistence:persistence-api:1.0.1 on the
central repo :) Which is probably not an easy thing to do...
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
11:04 a.m.
New subject: [Hibernate-JIRA] Commented: (ANN-712) Maven pom.xml should (in time) depend on javax.persistence:persistence-api instead of org.hibernate:ejb3-persistence
[
http://opensource.atlassian.com/projects/hibernate/browse/ANN-712?page=co...
]
Chris Bredesen commented on ANN-712:
------------------------------------
But the TCK and the APIs evolve over time to cope with TCK or Spec
bugs.
TCK isn't the issue here (at least I don't think). If the API ==
persistence-api-1.0.jar then how can that change over time?? It would become 1.0.1 or
some other version. Am I missing something here? What's in our copy that isn't
in the published JSR copy?
Maven pom.xml should (in time) depend on
javax.persistence:persistence-api instead of org.hibernate:ejb3-persistence
--------------------------------------------------------------------------------------------------------------------
Key: ANN-712
URL: http://opensource.atlassian.com/projects/hibernate/browse/ANN-712
Project: Hibernate Annotations
Issue Type: Task
Affects Versions: 3.3.1.GA
Reporter: Geoffrey De Smet
Priority: Minor
Now it's possible to get the same classes (for example javax.persistence.Entity)
twice in the classpath:
- one from persistence-api-1.0.jar
- and another from ejb3-persistence-1.0.1.GA.jar.
Mix that with classloaders and you get bugs like A extends B but an A is not instance of
B (because B.class is loaded twice from different jars). Otherwise you can get this too,
but if the groupId:artifactId match, maven deals with it.
In theory, replace:
<dependency>
<groupId>org.hibernate</groupId>
<artifactId>ejb3-persistence</artifactId>
<version>1.0.1.GA</version>
</dependency>
with:
<dependency>
<groupId>javax.persistence</groupId>
<artifactId>persistence-api</artifactId>
<version>1.0</version> <!-- FIXME -->
</dependency>
It looks like org.hibernate:ejb3-persistence is just a copy of
javax.persistence:persistence-api, but with a hot-fix applied to it:
probably because javax.persistence:persistence-api:1.0.1 isn't deployed to the
central repo yet.
So the JCP might need to be poked to put javax.persistence:persistence-api:1.0.1 on the
central repo :) Which is probably not an easy thing to do...
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
5:01 p.m.
New subject: [Hibernate-JIRA] Commented: (ANN-712) Maven pom.xml should (in time) depend on javax.persistence:persistence-api instead of org.hibernate:ejb3-persistence
[
http://opensource.atlassian.com/projects/hibernate/browse/ANN-712?page=co...
]
Emmanuel Bernard commented on ANN-712:
--------------------------------------
If you look at the history of ejb3-persistence and javax.persistence, they have not always
been aligned. And yes there were TCK issues, ask Shelly.
Think about it, what will we do the minute we start implementing JPA 2? Change the
dependency from javax.persistence to org.hibernate then move back once java.net catches
up? What if a TCK bug is blocking us, should we wait for weeks for java.net to catch up as
well?
Yes in a perfect pink world full of should have would have, we would use this package. But
reality is different.
Maven pom.xml should (in time) depend on
javax.persistence:persistence-api instead of org.hibernate:ejb3-persistence
--------------------------------------------------------------------------------------------------------------------
Key: ANN-712
URL: http://opensource.atlassian.com/projects/hibernate/browse/ANN-712
Project: Hibernate Annotations
Issue Type: Task
Affects Versions: 3.3.1.GA
Reporter: Geoffrey De Smet
Priority: Minor
Now it's possible to get the same classes (for example javax.persistence.Entity)
twice in the classpath:
- one from persistence-api-1.0.jar
- and another from ejb3-persistence-1.0.1.GA.jar.
Mix that with classloaders and you get bugs like A extends B but an A is not instance of
B (because B.class is loaded twice from different jars). Otherwise you can get this too,
but if the groupId:artifactId match, maven deals with it.
In theory, replace:
<dependency>
<groupId>org.hibernate</groupId>
<artifactId>ejb3-persistence</artifactId>
<version>1.0.1.GA</version>
</dependency>
with:
<dependency>
<groupId>javax.persistence</groupId>
<artifactId>persistence-api</artifactId>
<version>1.0</version> <!-- FIXME -->
</dependency>
It looks like org.hibernate:ejb3-persistence is just a copy of
javax.persistence:persistence-api, but with a hot-fix applied to it:
probably because javax.persistence:persistence-api:1.0.1 isn't deployed to the
central repo yet.
So the JCP might need to be poked to put javax.persistence:persistence-api:1.0.1 on the
central repo :) Which is probably not an easy thing to do...
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
2:01 a.m.
New subject: [Hibernate-JIRA] Commented: (ANN-712) Maven pom.xml should (in time) depend on javax.persistence:persistence-api instead of org.hibernate:ejb3-persistence
[
http://opensource.atlassian.com/projects/hibernate/browse/ANN-712?page=co...
]
Geoffrey De Smet commented on ANN-712:
--------------------------------------
Fair enough :)
Maybe a comment in the hibernate-annotations's pom and hibernate-entitymanager's
pom might make it clear for other users as well:
<!-- java.net's javax.persistence:persistence-api is not always up to date -->
Maven pom.xml should (in time) depend on
javax.persistence:persistence-api instead of org.hibernate:ejb3-persistence
--------------------------------------------------------------------------------------------------------------------
Key: ANN-712
URL: http://opensource.atlassian.com/projects/hibernate/browse/ANN-712
Project: Hibernate Annotations
Issue Type: Task
Affects Versions: 3.3.1.GA
Reporter: Geoffrey De Smet
Priority: Minor
Now it's possible to get the same classes (for example javax.persistence.Entity)
twice in the classpath:
- one from persistence-api-1.0.jar
- and another from ejb3-persistence-1.0.1.GA.jar.
Mix that with classloaders and you get bugs like A extends B but an A is not instance of
B (because B.class is loaded twice from different jars). Otherwise you can get this too,
but if the groupId:artifactId match, maven deals with it.
In theory, replace:
<dependency>
<groupId>org.hibernate</groupId>
<artifactId>ejb3-persistence</artifactId>
<version>1.0.1.GA</version>
</dependency>
with:
<dependency>
<groupId>javax.persistence</groupId>
<artifactId>persistence-api</artifactId>
<version>1.0</version> <!-- FIXME -->
</dependency>
It looks like org.hibernate:ejb3-persistence is just a copy of
javax.persistence:persistence-api, but with a hot-fix applied to it:
probably because javax.persistence:persistence-api:1.0.1 isn't deployed to the
central repo yet.
So the JCP might need to be poked to put javax.persistence:persistence-api:1.0.1 on the
central repo :) Which is probably not an easy thing to do...
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
6126
days inactive
6128
days old
hibernate-issues@lists.jboss.org
4 comments
3 participants
participants (3)
-
Chris Bredesen (JIRA) -
Emmanuel Bernard (JIRA)
-
Geoffrey De Smet (JIRA)