With regard to the naming strategy, I do not believe that is accurate. Although I will say that tracking exactly what happens with naming stratagies on 4.2 and 4.3 is a PITA because of some changes made there that I refused to allow into master. Even 4.3 was not always compliant with the JPA spec in terms of implicit names for tables/columns. With 5.0 you actually can achieve compliance by specifying an alternate ImplicitNamingStrategy. By default we use org.hibernate.boot.model.naming.ImplicitNamingStrategyLegacyJpaImpl. You would want to specify org.hibernate.boot.model.naming.ImplicitNamingStrategyJpaCompliantImpl. Like I said, I am pretty confident that 5.0 is backwards compatible with 4.3 in terms of how it handled these names by default. Now if you supplied a custom naming strategy....
Well this issue is (currently) about schema creation. But your exception is in regards to schema migration. Different beasts. So yes, should be different issues. Or we can rename/repurpose this one.
As far as me checking out Spring and running your tests... I'd obviously prefer a simplified test case, for many many reasons (http://sscce.org/) but if that's what you got then I will get to that when I have time.
|