Hi Tom,
While we are waiting for others' comments, I just want to say a couple
more reasons in favor of the change:
1) We already support features that make persistence.xml non-portable.
e.g. we don't require MappedSuperclass or Embeddable class names to be
listed in persistence.xml. We support <jar-file> in Java SE environment
which is not required to be supported.
2) Our competitors are taking this feature for granted as the following
comment shows...
http://www.theserverside.com/news/thread.tss?thread_id=38082#193940.
Thanks again for getting back so quickly.
Sahoo
Tom Ware wrote:
> Hi Sahoo,
>
> I am somewhat undecided about whether this change is a good idea or not.
>
> You have done a good job at pointing out the benefits of this change
> in the bug. (https://glassfish.dev.java.net/issues/show_bug.cgi?id=1258)
>
> The drawback of the fix is in the issue of portability. By the spec:
>
> "To insure the portability of a Java SE application, it is necessary
> to explicitly list
> the managed persistence classes that are included in the persistence
> unit."
>
> As a result, with this change, by default, an application that makes
> use of this default is not portable. Before this fix, the user had to
> explicitly specify something that would make their application
> non-portable.
>
> I am not totally opposed to this change. I just think it merits some
> discussion. Does anyone have any comments?
>
> Assuming we come to a consensus about the desirability of the change,
> the code change is fine.
>
> -Tom
>
> Sanjeeb Kumar Sahoo wrote:
>
>> Hi Tom,
>>
>> Please review the fix for
>> https://glassfish.dev.java.net/issues/show_bug.cgi?id=1258
>>
>> In order to test this functionality, I have also changed one of the
>> persistence-units in entity-persistence-tests to not specify
>> exclude-unlisted-classes.
>>
>> Thanks,
>> Sahoo
>>
>>