>From a JPA standpoint, you'll need to either pass along the credential
info (or use a ThreadLical or some sort of context-specific state), or
use the underlying capabilities of your database to do per-user
authentication. For example, Oracle's OCI driver allows communication
of user authorization
Hi,
I think that you should be using the regular Java type name; are there
really underscores and no package in the class ne?
You can always specify a strategy for a particular field with the
@Strategy annotation as well.
-Patrick
On 12/3/07, Roy Smith <[EMAIL PROTECTED]> wrote:
> Thanks for th
Sounds like a bug to me. If you could create a test case and attach it
to a new JIRA issue, that'd be great.
-Patrick
On 12/3/07, Marc Siegel <[EMAIL PROTECTED]> wrote:
> Hi,
>
> 1. The subclasses have additional attributes, but do not modify these at
> all.
> 2. The mapped superclasses are enh
Hi,
That's a bug; OpenJPA is not translating exceptions thrown during
JTA-managed commits. Can you open a JIRA issue for it?
-Patrick
On 11/30/07, Ajay Aggarwal <[EMAIL PROTECTED]> wrote:
> Patrick,
>
> Here is the full stack trace.
>
> -Ajay
>
> occurred during commit of transaction Name=[
Hi,
The posted stack trace looks fine to me. The duplicate record is being
detected while writing to the database during commit, and this is
causing the tx to fail, hence the RollbackException.
If you changed your test to load all the Messages into memory in the
tx before calling persist() then y
... Unless your db has deferred constraints or uses optimistic locking
internally to maintain db locks, in which case the flushed tx may
still a
fail.
-Patrick
On 12/2/07, Craig L Russell <[EMAIL PROTECTED]> wrote:
> Hi Gul,
>
> If you execute a JPA method, the JPA exception thrown should be able
Hi,
I guess I don't totally understand the use case. Why do you want to
add a record to the result list? Semantically, the result list is the
list of all instances that match the query; adding an instance to the
query result list seems like it erodes this semantic meaning.
I am not surprised that
Hi,
What do the subclasses look like? Do they have setIsChanged() or
setSynchronizedAt() overridden definitions?
Are the mapped superclasses enhanced?
-Patrick
On 11/30/07, Marc Siegel <[EMAIL PROTECTED]> wrote:
> Hi Folks,
>
> I've noticed a certain bug that appears in some, but all of the cas
Hi,
I think ValueHandler will be the way to go. This will let you use the
new type as a field, as a member of a list, and as a key or value in a
map.
Additionally, you'll be able to use instances of the new type in queries.
There is a sample ValueHandler for Points that should map well to your n
Hi,
A couple of guesses:
- have you already modified anything in the transaction before calling
executeUpdate()? This could trigger OpenJPA to do the update in memory
in some situations, which would explain the SELECT.
- I think that the sadly-unlocalized message is caused by your attempt
to upd
Hi,
Do you have the full exception stack? That's just the outer exception.
-Patrick
On 11/29/07, Ajay Aggarwal <[EMAIL PROTECTED]> wrote:
> Well, I am seeing OptimisticException and not OptimisticLockException.
> What's the difference between the two? Below is my trace, if that's any
> hint as t
Yes; often, there will be a nested exception available from
Exception.getCause().
-Patrick
On 11/29/07, Gul Onural <[EMAIL PROTECTED]> wrote:
>
> When a database error occurs in the openjpa, I always get rollback
> exception. For example if there is an error
> because of a duplicate record creati
Hi,
What do you mean by this?Are you saying that you can't insert or
delete records in the list, or that you can't modify entries in the
list? I would expect the former regardless of using the enhancer.
-Patrick
On 11/29/07, Sebastian Souza <[EMAIL PROTECTED]> wrote:
>
> I cant modify any result
Hi,
The internal exception should be re-wrapped in an OptimisticLockException.
-Patrick
On 11/28/07, Adam Hardy <[EMAIL PROTECTED]> wrote:
> Ajay Aggarwal on 28/11/07 17:00, wrote:
> > Hi,
> >
> >
> >
> > I wanted to know why openJPA uses
> > "org.apache.openjpa.util.OptimisticException" and not
Hi,
I think that I made that change as part of the changes to handle
serialization of EntityManagers. However, issues with BrokerFactory
caching based on full properties, so I switched to just use the
configuration's id. So, it's probably fine to just revert that one
change.
-Patrick
On 11/27/07
Hi,
If the fix is checked into the 1.0.x branch already (check the svn
commit data in the jira issue), then the 1.0.2 snapshot build should
have the feature.
-Patrick
On 11/25/07, klaasjan elzinga <[EMAIL PROTECTED]> wrote:
> Judging from the jira comments you will have to wait for 1.0.2. The
>
Hi,
What you're looking for is a unidirectional one-to-one. Can you post
the source code for the second case in your email?
-Patrick
On 11/21/07, Gul Onural <[EMAIL PROTECTED]> wrote:
>
>
> I have a classA which has a one-to-one relationship to classB. The
> classB side of the relationship is re
Yes, only things marked @embeddable can be put into @Embedded fields.
As Craig points out, this is certainly a candidate for a better test case.
-Patrick
On 11/24/07, Sebastian Souza <[EMAIL PROTECTED]> wrote:
>
> I found a solution to my problem
>
> my embedded field was a type Data (a mapped s
Hi,
To run the enhancer via ant, you need to have the classes available in
the classpath used when the ant task is run. How are you invoking the
ant task?
-Patrick
On 11/24/07, gutostraube <[EMAIL PROTECTED]> wrote:
>
> Hi all!
>
> Every time I start the server the following message appears on t
Ah. Are you attempting to persist new instances with null blob values?
-Patrick
On 11/22/07, Luc1fer Hell <[EMAIL PROTECTED]> wrote:
> If i post the SQL), it would be about 30 mb).
> I think i found cause of problem in Postgre log file.
> It starts with:
> 2007-11-22 16:27:58 ERROR: null value i
Hi,
Can you post the SQL that caused this? You can turn this on by setting
the openjpa.Log property to SQL=TRACE.
-Patrick
On 11/22/07, Luc1fer Hell <[EMAIL PROTECTED]> wrote:
> Hello.
>
> I'm trying to implement data replication mechanism base on OpenJPA and
> WebServices.
> I almost solve the
Hi,
It sounds like your relations are lazy (the default for collections
and maps). What happens if you change them to be eager?
-Patrick
On 11/16/07, Nick Johnson <[EMAIL PROTECTED]> wrote:
> Just a thought here... is OpenJPA using this strategy because it's
> optimized to hope that objects will
If you are using JTA, your tx manager will take care of this for you.
That's the only spec-compliant way right now.
I'm sure that any decent implementation will have an apo for this, so
you might consider creating a class that you call to prform this and
other common extensions in a single locatio
They're both part of the JTA spec.
-Patrick
On 11/13/07, Adam Hardy <[EMAIL PROTECTED]> wrote:
> Patrick Linskey on 10/11/07 21:15, wrote:
> > Hi,
> >
> >> if openjpa.TransactionMode=local, we have to deal with the transactions
> in the
> >> data access layer. But this is not what I want to do -
Hi,
There are some cases where we know an obj is detached, and optimize
away the SELECT. I don't remember the exact cases right now, and don't
have access to the docs, so this is probably only of limited
assistance.
It'd be good to know more about your domain model, though.
-Patrick
On 9/18/07,
Regarding q2: em.merge() should do the trick.
Also, IIRC, your case 1 should always fail; maybe we should make it
throw instead since that's a pretty convoluted way to always return no
results.
-Patrick
On 9/17/07, frankca <[EMAIL PROTECTED]> wrote:
>
> Hi,
>
> I have simple query with Foo and F
Hi,
Yes, they must be managed instances. When using a persistent type in a
query, you are either doing object equality (in which case we must
have a managed instance to understand how to handle object identity)
or you are accessing persistent fields (in which case we must be
hooked into the object
Hi,
Can you post the full stack, including the nested exceptions?
-Patrick
On 9/17/07, Diaz Lara JuanManuel <[EMAIL PROTECTED]> wrote:
> The error is send afte 2,000 rows iterated when using OpenJPA + Derby
> Client Driver, but it works ok when using TopLink + Derby Client Driver,
> same driver
Yep, it's some sort of a classloader issue. Somehow, PersitenceCapable
is available in two different classloaders. I'll know more later today
/ tomorrow.
-Patrick
On 8/30/07, ben short <[EMAIL PROTECTED]> wrote:
> My test apps not rocket science so feel free to make a test case for
> it. But coul
Yes. Do you want the different EMFs to point to the same tables? If
so, and if you use data caching, OpenJPA will properly handle
intra-JVM cache invalidation.
-Patrick
On 8/22/07, Gavin Bong <[EMAIL PROTECTED]> wrote:
> Hi,
>
> Is it possible to have multiple EntityManagerFactories to point to t
30 matches
Mail list logo
