[ Post a follow-up to this message ]

Craig L Russell wrote:
> It seems that we added PersistenceUnitName to the
> PersistenceManagerFactory before completely understanding the usage,
> which primarily is to assign a name when configuring the PMF via
> jdoconfig.
>
> I now believe that the "PersistenceUnit" part of the property name is
> distracting, and the real property name should be simply "Name". That
> is, PersistenceManagerFactories can have a Name property. The javadoc
> and other documentation can call out that this name is the same as the
> PersistenceUnit name in JPA, but I don't think that there is value in
> calling it PersistenceUnitName.
>
> The implication is a change in the PersistenceManagerFactory API and
> the jdoconfig xsd.
>
> Any comments?
That sounds good.

I would like again to introduce the idea that we also add such a name
property to the
PersistenceManager and that we add a getPersistenceManagerByName(String
aName)
to the PersistenceManagerFactory.

cheers
Christian

--
Christian Ernst
Software Engineer

Tel: +49-40-60990 338
Fax: +49-40-60990 113
EMail: cernst@versant.com

Versant GmbH
Wiesenkamp 22b
22359 Hamburg
Germany

Think Outside the Grid!
http://www.versant.com

Versant GmbH is incorporated in Germany.
Company registration number: HRB 54723, Amtsgericht Hamburg.
Registered Office: Wiesenkamp 22b, 22359 Hamburg, Germany.
Geschäftsführer: Jochen Witte

[ Post a follow-up to this message ]

I agree.  There is no need to be so verbose, and the concept of a persistenc
e unit was and is fully embodied by JDO's PersistenceManagerFactory.  The on
ly new concept, IMHO, is to name it, and "name" is sufficient to capture thi
s.  Can you enter a JIRA fo
r this and I'll make a patch?

----- Original Message ----
From: Craig L Russell <Craig.Russell@sun.com>
To: JDO Expert Group <jdo-experts-ext@sun.com>; Apache JDO project <jdo-dev@
db.apache.org>
Sent: Tuesday, June 5, 2007 5:46:19 PM
Subject: Rename property PersistenceUnitName to Name


It seems that we added PersistenceUnitName to the
PersistenceManagerFactory before completely understanding the usage,
which primarily is to assign a name when configuring the PMF via
jdoconfig.

I now believe that the "PersistenceUnit" part of the property name is
distracting, and the real property name should be simply "Name". That
is, PersistenceManagerFactories can have a Name property. The javadoc
and other documentation can call out that this name is the same as
the PersistenceUnit name in JPA, but I don't think that there is
value in calling it PersistenceUnitName.

The implication is a change in the PersistenceManagerFactory API and
the jdoconfig xsd.

Any comments?

Craig

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!

[ Post a follow-up to this message ]

While I kind of see your point (but not fully yet), I don't think this is s=
omething that we should standardize, given that yours is the only implement=
ation that I've encountered with this funtionality.  Thoughts?  Can you exp=
ound on its use, please?=0A=0A----- Original Message ----=0AFrom: Christian=
Ernst <cernst@versant.com>=0ATo: Craig L Russell <Craig.Russell@sun.com>=
=0ACc: JDO Expert Group <jdo-experts-ext@sun.com>; Apache JDO project <jdo-=
dev@db.apache.org>=0ASent: Tuesday, June 5, 2007 8:50:00 PM=0ASubject: Re: =
Rename property PersistenceUnitName to Name=0A=0A=0ACraig L Russell wrote:=
=0A> It seems that we added PersistenceUnitName to the =0A> PersistenceMana=
gerFactory before completely understanding the usage, =0A> which primarily =
is to assign a name when configuring the PMF via =0A> jdoconfig.=0A>=0A> I =
now believe that the "PersistenceUnit" part of the property name is =0A> di=
stracting, and the real property name should be simply "Name". That =0A> is=
, PersistenceManagerFactories can have a Name property. The javadoc =0A> an=
d other documentation can call out that this name is the same as the =0A> P=
ersistenceUnit name in JPA, but I don't think that there is value in =0A> c=
alling it PersistenceUnitName.=0A>=0A> The implication is a change in the P=
ersistenceManagerFactory API and =0A> the jdoconfig xsd.=0A>=0A> Any commen=
ts?=0AThat sounds good.=0A=0AI would like again to introduce the idea that =
we also add such a name =0Aproperty to the=0APersistenceManager and that we=
add a getPersistenceManagerByName(String =0AaName)=0Ato the PersistenceMan=
agerFactory.=0A=0Acheers=0AChristian=0A=0A-- =0AChristian Ernst=0ASoftware =
Engineer=0A=0ATel: +49-40-60990 338=0AFax: +49-40-60990 113=0AEMail: cernst=
@versant.com=0A=0AVersant GmbH=0AWiesenkamp 22b=0A22359 Hamburg=0AGermany=
=0A=0AThink Outside the Grid!=0Ahttp://www.versant.com=0A=0AVersant GmbH is=
incorporated in Germany.=0ACompany registration number: HRB 54723, Amtsger=
icht Hamburg.=0ARegistered Office: Wiesenkamp 22b, 22359 Hamburg, Germany.=
=0AGesch=E4ftsf=FChrer: Jochen Witte

[ Post a follow-up to this message ]

> I now believe that the "PersistenceUnit" part of the property name is
> distracting, and the real property name should be simply "Name". That
> is, PersistenceManagerFactories can have a Name property. The javadoc
> and other documentation can call out that this name is the same as
> the PersistenceUnit name in JPA, but I don't think that there is
> value in calling it PersistenceUnitName.

So now we would only have the "named PMF" specified via a PMF property to=20
support and implementations wouldn't be required to support reading of JPA=
=20
"persistence.xml" ?=20
I thought the general idea was to have more read-across between JPA and JDO=
-=20
ok having a "named PMF" does have its equivalent in JPA clearly, but users=
=20
can't then just take an existing "persistence.xml" file and a=20
<persistence-unit> contained within it and use it under JDO. Or maybe I=20
misunderstand your proposal


=2D-=20
Andy =A0(Java Persistent Objects - http://www.jpox.org)

[ Post a follow-up to this message ]

On 6 June  2007 07:04, Andy Jefferson wrote: 
>
> So now we would only have the "named PMF" specified via a PMF property to
> support and implementations wouldn't be required to support reading of JPA
> "persistence.xml" ?
> I thought the general idea was to have more read-across between JPA and J=
DO
> - ok having a "named PMF" does have its equivalent in JPA clearly, but
> users can't then just take an existing "persistence.xml" file and a
> <persistence-unit> contained within it and use it under JDO. Or maybe I
> misunderstand your proposal

Or maybe you intend that an implementation supports both, and that the same=
=20
name is used for both. But in that case how does the implementation know=20
which it should go for ? What if a named PMF *and* a named "persistence-uni=
t"=20
exists for the supplied name - does "jdo.xml" take precedence over=20
"persistence.xml", or vice-versa?


=2D-=20
Andy =A0(Java Persistent Objects - http://www.jpox.org)

[ Post a follow-up to this message ]

[ Post a follow-up to this message ]

Hi Craig,

> This part of the use case was not in mind when I wrote the change
> request. The user asks for PMF by name "MyPMF" but then the
> PersistenceUnitName in MyPMF refers to a PersistenceUnit in the
> persistence.xml.
>
> Is that what JPOX does with the PersistenceUnitName today?

JPOX current behaviour :
If the user specifies a set of PMF properties and they include a=20
"persistenceUnitName" then JPOX will parse any persistence.xml (in the JPA=
=20
location(s)) for a persistence-unit of that name. With the persistence-unit=
=20
info it will take the properties specified in the persistence-unit and appe=
nd=20
on any others passed in to the PMF construction (in the same way as it is=20
done for an EMF with JPA). It will also internally load up the MetaData for=
=20
all classes specified by the persistence-unit (effectively removing the nee=
d=20
to discover metadata at runtime).
 
>
> Good question. Perhaps there is room for both PersistenceUnitName and
> Name properties? The Name is used to match the user's
> getPersistenceManagerFactory(String name) and the PersistenceUnitName
> is used to match the persistence-unit name attribute in persistence.xml?

I think that's what I would prefer since while they represent very similar=
=20
concepts maybe there are differences in behaviour of the implementation if=
=20
its a persistence-unit, or if its a named PMF.


Regards
=2D-=20
Andy =A0(Java Persistent Objects - http://www.jpox.org)

[ Post a follow-up to this message ]