openjpa-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Matthew T. Adams (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (OPENJPA-1744) Support JDO
Date Fri, 01 Apr 2011 14:23:05 GMT

    [ https://issues.apache.org/jira/browse/OPENJPA-1744?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13014572#comment-13014572
] 

Matthew T. Adams commented on OPENJPA-1744:
-------------------------------------------

@Todd, if Cassandra were to develop a query language, it could be supported thanks to JDO's
extensible query language design.  JDO allows the user to pass string-based queries in the
underlying datastore's native query language.

I feel that JDO is better suited for nonrelational back ends than JPA is, primarily due to
the relational-centric view JPA takes due to its charter of only targeting relational databases.
 JDO supports both relational and nonrelational datastores by design.  I feel that it's more
than just a matter of preference.  By choosing JDO, the architect is future-proofing (and
"present-proofing", I suppose) his application so that it can be used against wildly varying
datastores, with only configuration changes to deal with.  That's very powerful stuff, IMHO.

Andy Jefferson over at DataNucleus has been including support for many of the NoSQL datastores,
and finds that JDO is perfectly suited to support them.

@Michael, there are almost no "collisions" between JDO & JPA.  If you could clarify what
you mean by collision, I could respond better.

The current JPA (2.0 as of this writing) is almost a complete functional subset of JDO's functionality
(currently at version 3.0), and the work we're doing in JPA 2.1 is beginning to include features
that have been in JDO for a long time:  fetch plans.  The first discussion in JPA 2.1 culminated
in the use of downcasting in JPQL, a feature that had been in JDO since 1.0.

I encourage you to support JDO for your users.  After all, it's how your product got its start.
 Perhaps the JDO implementation code could be obtained from Oracle and updated to work against
the OpenJPA kernel.  It would at least be worth investigating.

> Support JDO
> -----------
>
>                 Key: OPENJPA-1744
>                 URL: https://issues.apache.org/jira/browse/OPENJPA-1744
>             Project: OpenJPA
>          Issue Type: Improvement
>            Reporter: Matthew T. Adams
>            Assignee: Michael Dick
>
> OpenJPA came from Kodo, which was a JDO implementation at the time the code was forked.
 This request is to bring compliance with the latest JDO specification back to the implementation
so that it can be used with nonrelational backends like Google BigTable, Hadoop, etc.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message