Yes I saw your email previously. I think my bug is the same as yours - so
it seems at least.
But I am not too thrill about having to change my model layer at this point
- I am also using kodo jpa (the paid version from BEA) and it works with
it.
So my choices are to either keep my design and everything already done and
use an old openjpa version (0.9.6) or chnage the design and move to the
latest openjpa.
I guess I am hoping that someone from openjpa commiters could fix this bug
once and for all. And then our company could use openjpa as a standard for
OR mapping. But until then the tool is not ready yet for production. I mean
who knows what other problem might be in the code base.
The kicker would be if there is some new compelling features in the latest
openjpa - then having to chnage our design might be more acceptable since
we would win on some other level.
In particular I am struggling with the poor L2 caching support in the tool
(0.9.6) - if the latest openjpa (1.1.0) made improvement in this regard
then I would consider that a compelling reason to upgrade.
As a side note - leaving caching out of the spec was an amazing mistake in
my opinion. At least the interface should have been covered to improve
compatibility.
In any case - I am thinking about changing our design so that openjpa 1.0.2
would work. I am just not convinced it is worth it yet... still thinking
about it.
Frederic
iain
<[EMAIL PROTECTED]
t.com.au> To
[email protected]
cc
03/27/2008 03:32
PM Subject
Re: Issue with enhancer
Please respond to
[EMAIL PROTECTED]
che.org
Frederic (and all),
I reported a problem last week with enhancer and mappedsuperclass, and
unfortunately nobody was able to help.
I detailed the relevant class relationships there, and I suspect it is
to do with cycle detection (or not!). Of course I don't have cycles in
the inheritance hierarchy, but the subclasses hold references to other
classes that hold references to classes... that form a cycle.
As soon as I removed inheritance from the model (all the other class
reference remained), everything was fine.
Cheers, Iain
[EMAIL PROTECTED] wrote:
> I have 99 classes being enhanced. Their relationship is not too crazy -
> rather simple and conventional.
> All classes have a base class - therefore they all have an entity class
and
> a mappedsuperclass class.
> So in total 198 classes - 99 entities and 99 mappedsuperclass.
>
> I am precising this because someone else reported the same issue using
> mappedsuperclass as well.
>
> Frederic
>
> PS: Of course I can not send you the classes - the company would not like
> that very much... ;)
>
> But it works well with the kodo jpa version - the 0.9.6. So I am using it
> for now - it is just too bad becuase I would have liked to be able to use
> the open source openjpa version.
>
>
>
>
>
> "Kevin Sutter"
> <[EMAIL PROTECTED]
> om>
To
> [email protected]
>
cc
> 03/27/2008 12:27
> PM
Subject
> Re: Issue with enhancer
>
> Please respond to
> [EMAIL PROTECTED]
> che.org
>
>
>
>
>
>
> Hi Frederic,
> I'm sure there are many scenarios that can produce the dreaded OOM error.
> I'm assuming you were referring to Issue OPENJPA-258 (
> https://issues.apache.org/jira/browse/OPENJPA-258)? The fix for that
Issue
> did go into 1.0.2 and 1.1.0. So, it very well could be that you are
> surfacing another scenario that produces the OOM error.
>
> If you could post more specifics about your scenario and the results,
maybe
> we can at least determine if we're talking about the same problem or a
new
> one.
>
> Thanks,
> Kevin
>
> On Thu, Mar 27, 2008 at 12:47 PM, <[EMAIL PROTECTED]> wrote:
>
>
>> Hello -
>>
>> I was trying the enhance my classes with openjpa 1.0.2 as well as 1.1.0
>>
> and
>
>> I run into an OutOfMemoryError.
>>
>> These classes are the same one I am able to enhance using kodo JPA 0.9.6
>> from BEA.
>>
>> I am reporting this becuase I saw that the previous bug report on this
>> issue was marked as resolved - and I am afraid that it is not.
>>
>> Thank you very much,
>>
>> Frederic
>>
>>
>>
>
>
>
>
>
--
No virus found in this outgoing message.
Checked by AVG.
Version: 7.5.519 / Virus Database: 269.22.0/1342 - Release Date: 25/03/2008
10:26 AM
