db-jdo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Craig L Russell <Craig.Russ...@Sun.COM>
Subject Minutes: JDO TCK Conference Call Friday, Mar 27, 9 am PDT
Date Fri, 27 Mar 2009 19:08:18 GMT
Attendees: Michelle Caisse, Michael Bouschen, Craig Russell


1.  M2 repo error. Should be a straightforward fix. This file was  
autogenerated a while back but never updated for recent releases. AI  
Craig upload a new version of the metadata file to the apache repo  
which then gets synced with the central repo.

Now might be a good time to migrate to maven 2 for the project. Using  
maven 2, the metadata files would be generated correctly during  
packaging. Craig has already created a maven 2 pom for the api  
project; using maven 2 for the tck would require more work, probably  
creating a maven 2 plugin to handle the logic of running the iut and  
ri tests.

2. (JDO-623) Query cancel and timeout support (https://issues.apache.org/jira/browse/JDO-623

) Michael has added patches to the JIRA. AI everyone review Michael's  
comments and patches to the JIRA.

3. runtck.iut There have been issues running the RI as the iut. AI  
Michelle see if the runtck.iut still works with DataNucleus.

4. Other issues

Should the enhancer API be a required part of an implementation? It  
would make writing the tck easier, since the enhancer is currently the  
biggest difference among implementations.

Should the tck support the user's choice of maven 1.0.2 or 1.1? Seems  
like maven 1.1 works well enough to make it a requirement. AI Michael  
file a JIRA to update the README for the tck.

Matthew Adams sent an email last week regarding fetch notifications  
(see action item below). AI everyone read and react to the proposals.

There is renewed interest (e.g. cloud computing) in persistence APIs  
that don't have a fully ACID compliant, fully featured SQL back end,  
and yet the data is not simply key-value blob storage. This might be a  
good fit for JDO, assuming that we can adapt the data model and query  
model to make some features optional. Something like a profile  
approach (list a group of features that are required for the basic  
profile distinct from the features that the full profile needs).

Action Items from weeks past:

[Feb 13 2007] AI Craig update the spec and make clear the relationship  
between Query.cancel and multithreading.

[Feb 13 2007] AI Matthew discuss on email method by which an  
application might dynamically tune itself by tracking field access by  

[Nov 30 2007] AI Christiaan propose more details on Update/copy by  
query for post-JDO 2.1.

[May 25 2007]  AI everyone download the Grails demo from grails.org   
and check it out. Also look at Grails/Groovy ExpandoMetaClass that   
has the magic to avoid reflection and enhancement.

[May 25 2007] AI Matthew Adams prepare a proposal with just the basics  
of schema synchronization with jdo and orm metadata.

-- Michelle

Craig L Russell
Architect, Sun Java Enterprise System http://db.apache.org/jdo
408 276-5638 mailto:Craig.Russell@sun.com
P.S. A good JDO? O, Gasp!

View raw message