Folks,
The discussion on converter exceptions had me reconsidering duplicate key
exceptions.
Has anyone found it inconvenient that duplicate key exceptions are not
recoverable, in the sense that TX is marked for rollback?
> -----Original Message-----
> From: michael keith [mailto:michael.keith_at_oracle.com]
> Sent: Friday, 4 May 2012 6:59 a.m.
> To: jsr338-experts_at_jpa-spec.java.net
> Subject: [jsr338-experts] Re: converter errors
>
> Yes (if transaction exists).
>
> On 03/05/2012 1:32 PM, Linda DeMichiel wrote:
> > Our TCK engineer has pointed out to me that the spec doesn't
> > define what happens if a converter throws an exception.
> >
> > I would propose that the persistence provider wrap the
> exception in a
> > PersistenceException and mark the transaction for rollback.
> >
> > Other opinions?
> >
> > -Linda