db-jdo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Eric Samson" <eric.sam...@xcalia.com>
Subject RE: JDO2 Annotations
Date Thu, 13 Jul 2006 22:25:10 GMT
Hello Craig and all


Here is why we support option 3:


*	we think that annotations are nice for persistence but not for mapping information

	*	so annotation should remain as simple as possible and stay data source independent
	*	one goal of a persistence tool is to remove technical APIs to access data sources from
business applications source code, mapping annotations are a move backward in that case. Moreover,
they tend to make the source code unreadable as it is cluttered with dozens of interpreted
comment lines

*	it is important for users to have a kind of convergence between JDO and JPA in the future

	*	to that extent defining a set of completely different annotations is not desirable
	*	my guess is that ORM tools will have to support both JDO and JPA in the future so it is
better to implement only one set of annotations (maybe with additional JDO2 annotations whenever
	*	to me JDO2 is already a superset of JPA, so JDO annotations should be a superset of JPA


Best Regards,

....: Eric Samson, Founder & CTO, xcalia

Service your Data!


De : Craig.Russell@Sun.COM [mailto:Craig.Russell@Sun.COM] 
Envoyé : mercredi 12 juillet 2006 22:06
À : Apache JDO project
Cc : JDO Expert Group
Objet : Re: JDO2 Annotations


Hi Matthew,


With hindsight, do you think that this is the right solution? Any insights from your implementation
and usage experience whether it was a good or bad idea?




On Jul 12, 2006, at 1:02 PM, Matthew T. Adams wrote:

Xcalia supports option 3, reusing JSRs 220 & 250 annotations and defining additional ones
as necessary.


----- Original Message ----
From: Ilan Kirsh <kirsh@objectdb.com>
To: Craig L Russell <Craig.Russell@Sun.com>; jdo-dev@db.apache.org
Cc: JDO Expert Group <jdo-experts-ext@Sun.com>
Sent: Wednesday, July 12, 2006 12:06:55 PM
Subject: Re: JDO2 Annotations

After reviewing Andy's proposal and the relevant page on JPOX website

I vote for:

2. Define a complete set of annotations - based on JPOX current work

And IMO, this should be a required feature.


----- Original Message -----

From: "Craig L Russell" <Craig.Russell@sun.com <mailto:Craig.Russell@sun.com> >

To: <jdo-dev@db.apache.org <mailto:jdo-dev@db.apache.org> >

Cc: <jdo-experts-ext@sun.com <mailto:jdo-experts-ext@sun.com> >

Sent: Wednesday, July 12, 2006 7:36 PM

Subject: Re: JDO2 Annotations


> Javadogs,
> Please take a look at this proposal by Andy.
> High order bit: The expert group needs to make a decision whether to:
> 1. Not define any annotations, leaving it up to implementations to 
> decide what to do
> 2. Define a complete set of annotations
> 3. Track JSR 220 and JSR 250 annotations for persistence and mapping 
> and add only annotations for JDO that are not already covered by the 
> other annotation specifications
> Orthogonal to the above, we need to decide whether support of 
> annotations for implementations that support JDK 1.5 is required or 
> optional.
> Craig
> On Jul 4, 2006, at 11:48 PM, Andy Jefferson wrote:
>>> For information, you can find an initial (top-level) set of 
>>> suggested JDO2
>>> annotations at
>> http://jpox.cvs.sourceforge.net/jpox/JPOX/Plugins/Java5/src/java/ <http://jpox.cvs.sourceforge.net/jpox/JPOX/Plugins/Java5/src/java/>
>> org/jpox/annotations/
>> Since it's all quiet on annotations I'll provoke further :-)
>> The above link now shows an almost complete set of proposed JDO2 
>> annotations
>> (I changed "PersistentField" to "Field" to match the metadata 
>> element). The
>> advantages of matching annotations to metadata element are 
>> obvious ... people
>> don't need to relearn the terms they already know from metadata. 
>> The only
>> places where I haven't stuck to this are the top-level 
>> PersistenceCapable,
>> PersistenceAware annotations (since they make more sense, to me)
>> The only elements missing are Index, ForeignKey, Unique (i'll add 
>> these soon),
>> Interface, Property (which are low interest for me) and the 
>> recursive aspects
>> of FetchGroup and Embedded (which are restricted due to annotations 
>> JDK1.5
>> design). Needless to say that these are all supported in the latest 
>> nightly builds where you could actually use them interchangeably 
>> with your
>> metadata, or JPA annotations, or indeed your own set of annotations 
>> if you
>> feel like it.
>> Comments are welcome. Really.
>> -- 
>> Andy
> Craig Russell
> Architect, Sun Java Enterprise System http://java.sun.com/products/jdo <http://java.sun.com/products/jdo>

> 408 276-5638 mailto:Craig.Russell@sun.com <mailto:Craig.Russell@sun.com> 
> P.S. A good JDO? O, Gasp!



Craig Russell

Architect, Sun Java Enterprise System http://java.sun.com/products/jdo

408 276-5638 mailto:Craig.Russell@sun.com

P.S. A good JDO? O, Gasp!


  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message