[hibernate-dev] HHH-9329

Steve Ebersole steve at hibernate.org
Fri Sep 23 12:50:21 EDT 2016


Depends on the "fix" we all agree on.  Mainly whether that leads to any
non-compatible SPI changes.


On Fri, Sep 23, 2016 at 11:41 AM Christian Beikov <
christian.beikov at gmail.com> wrote:

> Ahh okay, I see. Sorry, I didn't fully read my previous mail and forgot I
> asked that question ^^
> Well then I'll try upgrading to 5.2 and hope for the best :)
>
> So are you considering merging that to 5.2 then?
>
>
>
> Am 23.09.2016 um 18:36 schrieb Steve Ebersole:
>
> Nope.  You asked:
>
> Are the problems with Hibernate 5.1+ and Infinispan fixed yet? I didn't
> consider upgrading yet because I read of some issues.
>
> Now its possibly I misread your reference to HIbernate+Infinispan problems
> to mean the only ones I know of. hence the HHH-10707 reference.  If you
> meant some other "Hibernate 5.1+ and Infinispan" problem, then I guess you
> could have been more specific ;)
>
>
>
> On Fri, Sep 23, 2016 at 11:33 AM Christian Beikov <
> christian.beikov at gmail.com> wrote:
>
>> Wrong thread?
>>
>>
>> Am 23.09.2016 um 17:53 schrieb Steve Ebersole:
>>
>> There are some conceptual mismatch problem that IMO stem from the L2C
>> SPI.  We are discussing that all as part of
>> https://hibernate.atlassian.net/browse/HHH-10707
>>
>> Whether that affects you really depends how you configure caching.  If
>> you try to reuse regions for different types of data (entity, collection,
>> etc) then it will affect you.  If you define different access strategies
>> for the same region then it will affect you.
>>
>>
>> On Fri, Sep 23, 2016 at 1:02 AM Christian Beikov <
>> christian.beikov at gmail.com> wrote:
>>
>>> Well to be fair, I already reported that this "regression" was
>>> introduced in 4.2.8 by placing the ON predicate on the target table join
>>> instead of the collection table join as was done before in 4.2.7.SP1.
>>> I am not using 4.x anymore but since Wildfly ships with Hibernate 5.0, I
>>> suppose that others could benefit from this too.
>>> Are the problems with Hibernate 5.1+ and Infinispan fixed yet? I didn't
>>> consider upgrading yet because I read of some issues.
>>>
>>>
>>> Regards,
>>> Christian
>>>
>>> Am 22.09.2016 um 22:10 schrieb Gail Badner:
>>> > Hi Christian,
>>> >
>>> > We are only backporting critical issues and regressions to 5.0 at this
>>> > point, so it can't be backported to 5.0. If no regressions caused by
>>> > this fix are reported in 5.2, I would consider backporting to 5.1.
>>> >
>>> > Regards,
>>> > Gail
>>> >
>>> > On Wed, Sep 21, 2016 at 12:14 AM, Christian Beikov
>>> > <christian.beikov at gmail.com <mailto:christian.beikov at gmail.com>>
>>> wrote:
>>> >
>>> >     Thanks, I know that it's a beauty ^^
>>> >     Hope this can get into all 5.x branches?
>>> >
>>> >     Regards,
>>> >     Christian
>>> >     Am 20.09.2016 um 23:05 schrieb Steve Ebersole:
>>> >     > I took a quick look.  I'd prefer to see better solution as we
>>> >     migrate
>>> >     > to SQM; but for 5.x, given how Hibernate generates SQL there, I
>>> >     am not
>>> >     > sure how else you would possibly do this
>>> >     >
>>> >     >
>>> >     > On Tue, Sep 20, 2016 at 12:54 PM Vlad Mihalcea
>>> >     > <mihalcea.vlad at gmail.com <mailto:mihalcea.vlad at gmail.com>
>>> >     <mailto:mihalcea.vlad at gmail.com <mailto:mihalcea.vlad at gmail.com>>>
>>> >     wrote:
>>> >     >
>>> >     >     Thanks,
>>> >     >
>>> >     >     I'm going to review it tomorrow.
>>> >     >
>>> >     >     Vlad
>>> >     >
>>> >     >     On Tue, Sep 20, 2016 at 8:38 PM, Christian Beikov <
>>> >     >christian.beikov at gmail.com <mailto:christian.beikov at gmail.com>
>>> >     <mailto:christian.beikov at gmail.com
>>> >     <mailto:christian.beikov at gmail.com>>> wrote:
>>> >     >
>>> >     >     > Hey again,
>>> >     >     >
>>> >     >     > I implemented the approach that I proposed in the issue
>>> and a
>>> >     >     test in
>>> >     >     > https://github.com/hibernate/hibernate-orm/pull/1561
>>> >     <https://github.com/hibernate/hibernate-orm/pull/1561>
>>> >     >     >
>>> >     >     > It detects left joins with join tables that use the target
>>> >     table
>>> >     >     alias.
>>> >     >     > The join table is replaced with a subquery and the WITH
>>> clause
>>> >     >     is moved
>>> >     >     > to the join of the subquery.
>>> >     >     >
>>> >     >     > Any comments?
>>> >     >     >
>>> >     >     > Regards,
>>> >     >     > Christian
>>> >     >     > _______________________________________________
>>> >     >     > hibernate-dev mailing list
>>> >     >     > hibernate-dev at lists.jboss.org
>>> >     <mailto:hibernate-dev at lists.jboss.org>
>>> >     <mailto:hibernate-dev at lists.jboss.org
>>> >     <mailto:hibernate-dev at lists.jboss.org>>
>>> >     >     >https://lists.jboss.org/mailman/listinfo/hibernate-dev
>>> >     <https://lists.jboss.org/mailman/listinfo/hibernate-dev>
>>> >     >     >
>>> >     >     _______________________________________________
>>> >     >     hibernate-dev mailing list
>>> >     > hibernate-dev at lists.jboss.org
>>> >     <mailto:hibernate-dev at lists.jboss.org>
>>> >     <mailto:hibernate-dev at lists.jboss.org
>>> >     <mailto:hibernate-dev at lists.jboss.org>>
>>> >     > https://lists.jboss.org/mailman/listinfo/hibernate-dev
>>> >     <https://lists.jboss.org/mailman/listinfo/hibernate-dev>
>>> >     >
>>> >
>>> >     _______________________________________________
>>> >     hibernate-dev mailing list
>>> >     hibernate-dev at lists.jboss.org <mailto:
>>> hibernate-dev at lists.jboss.org>
>>> >     https://lists.jboss.org/mailman/listinfo/hibernate-dev
>>> >     <https://lists.jboss.org/mailman/listinfo/hibernate-dev>
>>> >
>>> >
>>>
>>> _______________________________________________
>>> hibernate-dev mailing list
>>> hibernate-dev at lists.jboss.org
>>> https://lists.jboss.org/mailman/listinfo/hibernate-dev
>>>
>>
>>
>


More information about the hibernate-dev mailing list