I tried to explain in the original posting of the defect I am
heading towards similar annotations that use a in-database
, I understand that both the @CreationTimestamp and @UpdateTimestamp annotations generate a timestamp within the JVM. We are just starting to gain experience with Hibernate and thought that staring off with behavior that is part of the stock Hibernate distribution and then move towards our final desired behavior would be prudent rather than writing a lot of custom code that may not fit well with how Hibernate works or is intended to work. What I am trying to point out in this defect is that the behavior of the @UpdateTimestamp annotation and the @DynamicUpdate annotation do now work together. These annotations, when used separately, produced expected behavior. However when they are used in combination the logic behind @DynamicUpdate does not appear to detect a change in value that is generated by @UpdateTimestamp and in turn will not build SQL to update the associated column in the database. |