510898f57e
5.3.12.Final
2019-09-12 07:09:37 +01:00
b389c23780
5.3.12.Final
2019-09-11 20:52:57 -07:00
e812014ea9
HHH-13605 Upgrade the MariaDB JDBC driver to 2.2.4 in the matrix_mariadb task
...
It's already the version we use when running test -Pdb=mariadb, so let's
be consistent.
(cherry picked from commit ef87991fa3
2019-09-11 21:24:13 -05:00
c527ac15a5
HHH-13581 Disable tests involving MariaDB ConnectorJ's buggy ResultSet#getTime(int, Calendar) method
...
(cherry picked from commit debc5d37f8
2019-09-11 21:24:13 -05:00
a3c894e455
HHH-13580 Ignore LocalTimeTest and OffsetTimeTest's 'nativeWriteThenRead' test for MySQL
...
The returned time is right (otherwise #writeThenRead would fail), it's
just that a different day is returned, but it won't affect the LocalTime
representation manipulated by the user.
(cherry picked from commit 74d35b2d59
2019-09-11 21:14:54 -05:00
aef0405b5f
HHH-13580 Use safer ZoneId => TimeZone conversion in AbstractJavaTimeTypeTest
...
We were testing GMT instead of UTC-8 without even knowing it...
(cherry picked from commit 8fce51c765
2019-09-11 21:14:54 -05:00
58674a7ba7
HHH-13580 Fix a copy/paste error in OffsetTimeTest
...
(cherry picked from commit f9f6793232
2019-09-11 21:14:54 -05:00
b7c7bccc0b
HHH-13582 Ignore LocalDateTest for MySQL
...
MySQL ConnectorJ 8.x returns the wrong date
when the JVM default timezone is different from the server timezone:
https://bugs.mysql.com/bug.php?id=91112
(cherry picked from commit a580227569
2019-09-11 21:14:54 -05:00
61a23e0966
HHH-13582 Downgrade MySQL Connector/J to 8.0.12 to be consistent with WF18
2019-09-11 21:14:54 -05:00
52bdda7d40
HHH-13582 Upgrade MySQL Connector/J to 8.0.17
...
ConnectorJ 8 is the version used in WildFly integration tests.
ConnectorJ 5 is apparently no longer tested.
Note this solves most timezone-related issues we've been having.
(cherry picked from commit be7cc76556
2019-09-11 21:14:54 -05:00
5274f15ef5
HHH-13607 - Exception thrown while flushing uninitialized enhanced proxy with immutable natural ID
...
(cherry picked from commit 7ceaf3aaa4
2019-09-11 15:40:31 -05:00
c3286a2b56
HHH-13607 : Added tests
...
(cherry picked from commit 650b1dec6a
2019-09-11 15:40:31 -05:00
4b187995a9
HHH-13504 : Upgrade ByteBuddy to 1.9.11
...
(cherry picked from commit 450fdb75f5
2019-09-11 12:23:00 -07:00
59ad417498
HHH-12968 - Flush is not flushing inserts for inherited tables before a select within a transaction
...
Extract IdentityGenerator batch support validation logic
(cherry picked from commit f21c8c2927
2019-09-11 13:57:48 -05:00
fd186a1dd5
HHH-12968 - Flush is not flushing inserts for inherited tables before a select within a transaction
...
(cherry picked from commit 9dfdb2b471
2019-09-11 13:57:48 -05:00
af3231820a
HHH-13611 : Restore EntityMetamodel constructor to take SessionFactoryImplementor argument instead of PersisterCreationContext
2019-09-11 13:54:04 -05:00
996e60f685
HHH-13611 : Restore EntityMetamodel constructor to take SessionFactoryImplementor argument instead of PersisterCreationContext
2019-09-11 13:54:04 -05:00
8aa47b4e61
HHH-13592 Populate AutoFlushEvent#isFlushRequired with the correct value
...
(cherry picked from commit 5fc6012449
2019-09-11 13:25:21 -05:00
6c2be5224e
HHH-13592 Test AutoFlushEvent#isFlushRequired
2019-09-11 13:25:21 -05:00
793315267b
HHH-13616 Enable the hibernate-orm-modules test for JDK 11
2019-09-11 15:09:04 +01:00
dbbd2fcfb4
HHH-13428 No longer allow fetching snapshot dependencies from JBoss Nexus
2019-09-11 11:23:04 +02:00
1d8837d6f7
HHH-13428 Remove an unnecessary test dependency to woodstox
...
It has split packages and won't work when running tests in the module
path. But more importantly we just don't use it.
2019-09-11 11:23:04 +02:00
df20643bf8
HHH-13428 Upgrade to Microsoft SQL Server JDBC 7.2.1
2019-09-11 11:23:04 +02:00
ac43f3429f
HHH-13428 Remove more unused repositories
2019-09-11 11:23:04 +02:00
fcbd4dae31
HHH-13428 Use https in documentation too
2019-09-11 11:23:04 +02:00
21ad05ccb1
HHH-13428 Remove unused repository URLs
2019-09-11 11:23:04 +02:00
e7423b076f
Do not log test progress by default, but provide a build option and enable it for Travis
...
So that the build isn't too verbose when running it locally.
2019-09-11 10:26:35 +02:00
6e485989f1
Get rid of travis_wait in .travis.yml
...
... so that we get actionable logs when something fails.
2019-09-11 10:26:35 +02:00
5f5e029941
HHH-13415 Run Travis tests on JDK11 too
...
Use OpenJDK 11.0.3 instead of OracleJDK 11.0.2 in the Travis build,
because 11.0.2 just fails because of a bug in javadoc generation.
2019-09-11 10:26:35 +02:00
f2eeb0aed1
HHH-13415 Clean up the places where we fix fix javadoc generation for JDK11
2019-09-11 10:26:35 +02:00
433e12b371
HHH-13415 Add more comments where we disable OSGi tests with JDK11
...
Karaf or Pax Exam (don't know which) has serious issues with JDK 11.0.3.
2019-09-11 10:26:35 +02:00
ab527c0bd8
HHH-13421 Disable OSGi testing for JDK 11+
2019-09-11 10:26:35 +02:00
8a199f9e88
HHH-13419 Support building javadoc with JDK 11.0.3+
2019-09-11 10:26:35 +02:00
9251720f69
HHH-13275 Re-introduce usage of net.bytebuddy.experimental=true when testing on JDK > 11
2019-09-11 10:26:35 +02:00
2dbea0bbb2
HHH-13271 Restore compatibility of build script with Java 8 too
2019-09-11 10:26:35 +02:00
daf0a297df
HHH-13271 Javadoc build failures on JDK 12
2019-09-11 10:26:35 +02:00
38fd1be16d
HHH-13127 Update a few things in the metamodel generator documentation
2019-09-11 10:26:35 +02:00
d056e39cc1
HHH-13127 Document that JAXB jars are needed if using JDK 11
2019-09-11 10:26:35 +02:00
c344455df9
HHH-13128 Add missing jaxb-runtime dependency to metamodel generator
...
It's necessary for JDK 11 and we missed it when we added the
jaxb-runtime dependency to hibernate-core.
2019-09-11 10:26:35 +02:00
69fa165664
HHH-13043 Upgrade to JAXB 2.3.1
...
... which has a correctly formatted POM, and doesn't wreak havoc in
projects depending on us and using Maven + JDK11.
2019-09-11 10:26:35 +02:00
0e722b2fbe
HHH-13069 Go back to non-SSL links for our JBoss Nexus repositories
...
Using SSL repositories with JDK 11 ends up with the error below:
javax.net.ssl.SSLHandshakeException: extension (10) should not be presented in server_hello
This is a bug in JDK 11 that will apparently be fixed in the next update
(planned for January 2019).
Until then, we need to use non-SSL links.
For more information:
https://stackoverflow.com/questions/52016415/jdk-11-ssl-error-on-valid-certificate-working-in-previous-versions
https://bugs.openjdk.java.net/browse/JDK-8209965 - JDK 12 issue
https://bugs.openjdk.java.net/browse/JDK-8210005 - JDK 11 issue
2019-09-11 10:26:35 +02:00
8d67195aa8
HHH-13069 Update the links to JBoss Nexus to use the direct repository over https
2019-09-11 10:26:35 +02:00
f34fd25518
HHH-13022 Allow to fall back to the OSGi-provided TCCL during bootstrap
...
We previously erased that class loader, which is not a great idea, in
particular when we use libraries such as JAXB which rely on the TCCL, to
some extent.
Without this patch, we cannot make the OSGi integration work correctly
with Java 9 and above, because the JAXB APIs won't find the JAXB
runtime.
2019-09-11 10:26:35 +02:00
3e379edbf4
HHH-13022 Include karaf logs into test output files
2019-09-11 10:26:35 +02:00
25338d4471
HHH-13022 Re-enable OSGi tests when using JDK9/JDK10/JDK11
2019-09-11 10:26:35 +02:00
6cdbf13229
HHH-13022 Upgrade to Apache Karaf 4.2.1
2019-09-11 10:26:35 +02:00
0911dd384e
HHH-13022 Add missing versions to internal OSGi feature dependencies
2019-09-11 10:26:35 +02:00
625f341a56
HHH-13022 Fix indentation in hibernate-osgi.gradle
2019-09-11 10:26:35 +02:00
acc7934173
HHH-12990 Remove unnecessary dependencies to JSR-250 when generating JAXB classes
...
It seems these dependencies were not necessary at all, even before the
previous patches.
2019-09-11 10:26:35 +02:00
0f54fcc484
HHH-12990 Remove now unnecessary dependencies to JSR-250 when using jpamodelgen
...
The previous commit ensured we don't need them anymore.
2019-09-11 10:26:35 +02:00
Andrea Boriero
Gail Badner
Yoann Rodière
Steve Ebersole
Vlad Mihalcea
Javad Alimohammadi
Sanne Grinovero
Guillaume Smet