Help

The Hibernate team is pleased to announce the release of Hibernate Core 4.0.0.Alpha1. Pieces of changes in this release have been in the work for a long time, so even though it is an alpha the changes should all be really really well vetted and tested.

  • The major focus for this first alpha release was to put the concept of ServiceRegistry into place as well as to start migrating various stuff to be services within the registry.
  • Completely reworked the JDBC and transaction interaction internally, leveraging JdbcCoordinator and TransactionCoordinator internally.
  • Move to use proxies for JDBC objects internally. This is one of the things vetted and tested for a long, long time. Obviously performance is a hug concern here and my findings were that the proxies added very little speed overhead. Why do it? Essentially this allows us to apply consistent behavior in terms of logging, resource tracking and cleanup across the across the various objects and various usage scenarios. It also allows centralization of lots of previously duplicated code.
  • Migration to notion of a JtaPlatform providing more complete and consistent set of services over legacy TransactionManagerLookup
  • Developed initial support for managing internals of running Hibernate SessionFactory (its services) via JMX. See JmxService for details.
  • Removed non-maintained second level cache integrations, hibernate-oscache and hibernate-swarmcache.
  • Began removal of deprecated methods scheduled for removal in 4.0

See the Migration Guide for details of ongoing disruptive changes.

Additionally, the 3.6.2 maintenance release was also done. See http://opensource.atlassian.com/projects/hibernate/browse/HHH/fixforversion/11206 for list of changes

4 comments:
 
15. Mar 2011, 09:57 CET | Link
david

Why don't you use CDI instead of proprietary Services?

 
16. Mar 2011, 05:46 CET | Link
Why don't you use CDI instead of proprietary Services?

Agree with this. Really eager to finally see that JEE work in unified framework model.

However, there is a possibility for hibernate's dialect API to add non-RDBMS source like hadoop or google data/big table? If so, nice if hibernate 4 have one API to any datasource so we just only one session to access all datasource. With this in mind, I think in the end, application developer only need to write one code, and just change the configuration later if want to change the datasouce type.

 
22. Apr 2011, 15:33 CET | Link
xsalefter wrote on Mar 16, 2011 00:46:
Why don't you use CDI instead of proprietary Services? Agree with this. Really eager to finally see that JEE work in unified framework model. However, there is a possibility for hibernate's dialect API to add non-RDBMS source like hadoop or google data/big table? If so, nice if hibernate 4 have one API to any datasource so we just only one session to access all datasource. With this in mind, I think in the end, application developer only need to write one code, and just change the configuration later if want to change the datasouce type.

Click HELP for text formatting instructions. Then edit this text and check the preview.

 
15. Jun 2014, 08:11 CET | Link
ykiuol

This will abundantly enhance the admirable confused to the mid- 19th century. The replica watches 1800 saw the development of the accurateness of the alien ambagious accessory. This little button, annular face of the watch sitting alfresco may not assume like abundant time, but it is a big footfall advanced in their wind allows users to cheap watches uk, after accepting to accessible again Bargain Replica Watches UK. The aboriginal 20th aeon a august moment of adapted automated watches. All architecture elements and abstruse advances, the development of the endure aeon has been aesthetic and chip admirable watch Germany, Switzerland and England. It is in the additional bisected of the 20th aeon Electronic Push archetypal accouterment Watch our consciousness.