Upgrade to latest DataNucleus.

[GoogleCodeExporter]

Upgrade to DataNucleus 2.0

Original issue reported on code.google.com by max%oizo.biz@gtempaccount.com on 30 Jul 2009 at 4:17

• Blocking: #142, #178, #181

[GoogleCodeExporter]

2.0 isn't even final yet.  Is there a particular feature you need?

Original comment by max.r...@gmail.com on 3 Aug 2009 at 6:39

[GoogleCodeExporter]

Original comment by max.r...@gmail.com on 13 Oct 2009 at 5:22

• Changed state: Accepted

[GoogleCodeExporter]

Notes for migration, for things that may or may not be of relevance :-
1. There is a persistence property "datanucleus.singletonPMFForName" which you 
can
set to "true" to get simple handling to re-hand-out a PMF with a particular 
name. No
exception is thrown on subsequent attempts, just hands out the same one as 
already
exists. Presumably means you can delete your handling for singletons

2. You can override "org.datanucleus.java_type" entries by adding the 
override="true"
property to your setting. Obviously this makes you deviate from the JDO/JPA 
specs;
all DN defaults are strict specification compliant, as they should be.

3. persist/delete non-tx operations are atomic, hence committed straight away. 
In 1.1
the commit was only done on the subsequent update.

4. Many things are changed on the generic compile for other situations, in 
particular
JPQL (which is now close to passing the JPA TCK using the generic compiler based
process). This may mean you need to re-adjust your expectations (or maybe not).

5. L2 cache is now enabled. You may want to direct it to your javax.cache impl, 
and
include the datanucleus-cache plugin in the distribution.

Original comment by googleco...@yahoo.co.uk on 29 Oct 2009 at 5:20

[GoogleCodeExporter]

6. The StoreManager can register a listener for loading of metadata for a class.
Means that a datastore plugin can perform its own validity checking of metadata
appropriate for its support. So you can rip out any hacks to utilise your
MetaDataValidator, and just register a listener.

Original comment by googleco...@yahoo.co.uk on 1 Nov 2009 at 10:51

[GoogleCodeExporter]

7. StorePersistenceHandler now has batchStart/batchEnd for persist/delete bulk
operations (makePersistentAll, deletePersistentAll) allowing you to easily 
demarcate
batches and use if your underlying datastore needs to work like that. It also 
has
locateObjects/findObjects methods so that the datastore can do a bulk find if it
supports that.

Original comment by googleco...@yahoo.co.uk on 3 Nov 2009 at 2:25

[GoogleCodeExporter]

> 4. in particular JPQL (which is now close to passing the JPA TCK using the 
generic
compiler

Now does pass the JPA TCK (and also includes compilation of some JPA2 
constructs), so
is considered complete.

Original comment by googleco...@yahoo.co.uk on 8 Dec 2009 at 8:28

[GoogleCodeExporter]

I have a particular feature (actually bug fix) that I need.  The current JPA
implementation breaks when enriching an entity class that contains a method 
with the
signature of get(String) even if it has a @Transient annotation.  Unfortunately 
I
need it to have this signature because I am implementing an external interface.

The datanucleus forum thread can be seen here:
http://www.jpox.org/servlet/forum/viewthread_thread,5954_lastpage,yes#31826

I would really appreciate getting an update!

Thanks

Original comment by joe...@gmail.com on 2 Feb 2010 at 8:39

[GoogleCodeExporter]

DataNucleus 2.1 M1 is now released. This moves to using JPA2 final jar. It also
includes much internal restructuring in terms of API components. Also includes 
some
further improvements to generic query compilation.

What is the timescale for GAE/J's plugin working against DN 2+ ?

Original comment by googleco...@yahoo.co.uk on 3 Apr 2010 at 9:52

[GoogleCodeExporter]

I spent a few hours getting the GAE/J plugin compiling against DN2 but there 
are still 
quite a few test failures that need to be resolved.  Unfortunately there are 
other 
efforts absorbing all my time at the moment so I don't have a specific timeline.

Original comment by max.r...@gmail.com on 25 Apr 2010 at 9:23

[GoogleCodeExporter]

be aware that DN 2.1 is planned for full release June 2010

Original comment by googleco...@yahoo.co.uk on 26 Apr 2010 at 2:14

[GoogleCodeExporter]

DN 2.1 is now released ... fwiw

Original comment by googleco...@yahoo.co.uk on 10 Jun 2010 at 10:42

[GoogleCodeExporter]

Who (at google) cares weather DN 2.1 is released or not. !!

Original comment by sramana...@gmail.com on 17 Jun 2010 at 6:35

[GoogleCodeExporter]

Please support the latest stable Datanucleus version!

Original comment by george.m...@gmail.com on 8 Nov 2010 at 11:59

[GoogleCodeExporter]

http://www.jpox.org/servlet/forum/viewthread_thread,5954_lastpage,yes#31826

Original comment by miuler on 10 Jan 2011 at 9:55

[GoogleCodeExporter]

Executing a JDO query created using PersistenceManager.newNamedQuery fails when 
using a manager obtained through 
PersistenceManagerFactory.getPersistanceManageProxy():

javax.jdo.JDOException: PrimaryExpression should be a VariableExpression with 
name [Name_Of_Query]

Fixed in DataNucleus 2.0.

I am just adding it here hoping google search will pick this up for someone 
else experiencing the problem.

Original comment by arnold.m...@gmail.com on 8 Feb 2011 at 7:09

[GoogleCodeExporter]

DN 2.2 is released.  Further, there are no plans for further releases of DN 1.1.

My application wants to use JDO to access a remote data store (a Google 
Spreadsheets feed) using Data Nucleus' pluggable Store implementation.  I would 
rather write that store implementation for DN 2.0, not DN 1.x.

Original comment by ros.bull...@gmail.com on 4 Apr 2011 at 12:26

[GoogleCodeExporter]

FWIW. Best write your Google Spreadsheets store implementation for DN 3.0 since 
it significantly changes APIs.

Original comment by googleco...@yahoo.co.uk on 10 Apr 2011 at 6:16

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Please count my vote as a vote for DN 3.0.  It would be helpful to me to know 
when/if Google planned to upgrade DN, so that I could appropriately target my 
Store implementation efforts.

Original comment by dtbull...@gmail.com on 15 Apr 2011 at 3:29

[GoogleCodeExporter]

Any news after Google IO 2011?

Original comment by ritterch...@gmail.com on 12 May 2011 at 7:52

[GoogleCodeExporter]

Hey Matt,

Is it possible to get Data Nucleus upgraded to 2.1 or 3.0.

- Chris

P.S. Please... pretty please.

Original comment by ritterch...@gmail.com on 16 May 2011 at 2:49

[GoogleCodeExporter]

We're going to invest a little time to see how feasible it would be for us to 
start contributing to the Google Datastore implementation for DataNucleus. I 
posted a note on the DataNucleus Forum 
<http://www.datanucleus.org/servlet/forum/viewthread_thread,6715>. Ikai Lan 
said (in Appengine chat) that he'd check with Max Ross on what it might take to 
add a new committer.

If we can't find anyone else to collaborate with on this, we'll probably move 
to Objectify. It doesn't seem prudent to depend on Google to move this 
technology forward.

Original comment by jwbito on 19 May 2011 at 4:36

[GoogleCodeExporter]

Hi jwbito,

your suggestion is great! We are currently also thinking about upgrading the 
datanucleus-appengine plug-in in our company. The reason for doing so, is not 
to obtain better/newer features, but to use it in combination with another DN 
plug-in (currently under development), which is compatible with DN 3.X only. In 
this sense, this out-dated appengine plug-in is a clear show-stopper.

Next week, we will evaluate how feasible it would be to provide an update. Of 
course, it would be great to have partners or even a community, if we decide to 
start this project. 

Best regards
Jan

Original comment by jan.morl...@googlemail.com on 20 May 2011 at 2:14

[GoogleCodeExporter]

I'd certainly be happy to help with an up-to-date (and better?) DataNucleus 
plugin for Google Datastore - probably in the field of documentation.

Although objectify-appengine seems cute, the benefits of having a JDO/JPA 
interface for persistence include:

a) developer familiarity with existing standards;
b) access to other data-stores using the same interface;
c) an exit-strategy away from AppEngine in the future.

So this plugin certainly needs to be developed further.  It would be better if 
Google listened to their users and did it themselves though.  The dog-food is 
not tasting as good as it might.

Original comment by dtbull...@gmail.com on 21 May 2011 at 6:10

[GoogleCodeExporter]

Perhaps discussion of a community effort to develop a GAE datastore plugin for 
DN 3 should be held on the DataNucleus forum, at least until Max lets us know 
whether it's feasible to add committers on this project. (There are three 
others listed as committer/contributor, but only Max has made a change in the 
last 2 years.)

My company still needs to look at how feasible it would be to port between 
Objectify and JDO; that said, benefit (c) is the main reason that we like JDO.

It was heartening to see that Max updated the trunk with some contributed 
changes and made mention of a 1.1.0 version of the plugin. I'm not sure how the 
contribution was made, because I didn't find it attached to an issue. 

It would be great if there were a way to amplify Max's efforts and get his 
experience/expertise applied to a plugin for a modern version of DataNucleus. 
Maybe this will be the beginning of a community that makes a great contribution 
to GAE for Java.

Please consider joining the forum thread: 
http://www.datanucleus.org/servlet/forum/viewthread_thread,6715

Original comment by jwbito on 4 Jun 2011 at 3:57

[GoogleCodeExporter]

I'm really happy to see all this interest in moving the plugin forward!  We 
(Google) are planning a new major version of the plugin that will include an 
upgrade to the latest and greatest DN.  I'll update this bug as soon as I have 
more information.

Thanks!
Max

Original comment by max.r...@gmail.com on 6 Jun 2011 at 6:08

[GoogleCodeExporter]

Alright-y, then! It's good to hear from you, Max! If there's some useful 
community contribution to be made, please let us know.

Original comment by jwbito on 6 Jun 2011 at 8:08

[GoogleCodeExporter]

I'm always happy to accept patches for anything with a valid bug report, but 
please check with me first before starting on anything to make sure you're not 
stepping on anyone's toes.

Original comment by max.r...@gmail.com on 6 Jun 2011 at 8:16

[GoogleCodeExporter]

Original comment by max.r...@gmail.com on 6 Jun 2011 at 8:16

• Changed title: Upgrade to latest DataNucleus.
• Added labels: TargetRelease-2.0

[GoogleCodeExporter]

Issue 204 has been merged into this issue.

Original comment by max.r...@gmail.com on 6 Jun 2011 at 8:18

[GoogleCodeExporter]

Issue 226 has been merged into this issue.

Original comment by googleco...@yahoo.co.uk on 26 Jun 2011 at 6:02

[GoogleCodeExporter]

Issue 221 has been merged into this issue.

Original comment by googleco...@yahoo.co.uk on 26 Jun 2011 at 6:03

[GoogleCodeExporter]

Issue 124 has been merged into this issue.

Original comment by googleco...@yahoo.co.uk on 27 Jun 2011 at 4:29

[GoogleCodeExporter]

DataNucleus 3.0 is, to all intents and purposes, available for use with SVN 
trunk of this plugin. Other issues (with the plugin) remain but they are down 
to the basic fragility of the design of this plugin, and will be tackled in 
separate issues.

Original comment by googleco...@yahoo.co.uk on 10 Aug 2011 at 8:59

• Changed state: Fixed
• Added labels: Priority-High, Type-Enhancement
• Removed labels: Priority-Medium, Type-Defect

[GoogleCodeExporter]

Would you say the version present in the trunk is more or less fragile than
version 1.0.9?

Original comment by thegameo...@gmail.com on 22 Aug 2011 at 5:05

[GoogleCodeExporter]

We are Islamic Law Loans investor groups, we are projects financers based on 
Islamic Law. We as Islamic law financer lenders without interest payments, as a 
relationship that favors the lender, we obey the Islamic law, and we are not 
after charges interest at the expense of the borrower. Because Islamic law 
views money as a measuring tool for value and not an "asset" in itself, it 
requires that one should not be able to receive income from money (for example, 
interest or anything that has the genus of money. Our Islamic Law Loans 
investor groups seek only Islamic Capital Venture Fund seekers, with a 
registered company, Muslim community as a whole. While there are a number of 
Muslim countries that have adopted Western style legal systems â?? but that 
doesnâ??t mean Sharia doesnâ??t exist in those places. If you are a Muslim 
and you donâ??t belong to a Muslim country you would have to provide us with a 
letter from your Islamic Judicial Council, that you are a Muslim under it 
community, We are approved by Islamic Judicial Council. We give away Islamic 
Law Loans for all Protects types based on its legitimacy; our funding capital 
is an unlimited asset of over Euros â?¬500.000.000 Five Hundred Million Euros 
Candidates who wish to participate in the Process must submit interest by 
contacting out Admin Customer Care. Email:-: islamiclaws.loans@gncn.net

Original comment by info.Isl...@gmail.com on 6 Sep 2011 at 3:09

Attachments:

• [Islamic Logo.gif](https://storage.googleapis.com/google-code-attachments/datanucleus-appengine/issue-103/comment-36/Islamic Logo.gif)