Hibernate version: version 3.2 cr2

MySQL 4.1.19 + MySQL JDBC 3.1.12 driver :

I was trying to solve a problem with Hibernate + boolean mapped to char(1) column. I was getting an unexpected error because of a bug in the MySQL 3.1.12 driver. ResultSet.getBoolean(int index) wasn't clearing the internal null flag from the previous call to getString(int index).

The MySQL jdbc driver bug is fixed in the latest dev build.

I have a simple Hibernate unit test that recreates the problem.

Is there interest in checking in a Hibernate unit test that recreates a bug that exists in the MySQL JDBC driver?

Scott

And why do you think Hibernate developers should be concerned about bugs in MySQL drivers?

The purpose of this test would be to help ensure that the integration of Hibernate + MySQL is correct. It is currently not correct with regard to ResultSet.wasNull() being called after a call to ResultSet.getBoolean().

I think it would help Hibernate developers to have "integration tests" that verify that a backend database is working correctly with Hibernate. So, I think it is a good idea to check it in.

IMO the integration _is_ correct and it is the bug in the some party driver that prevents the tandem from working correctly. I am afraid that if they will follow the logic they would need to test for JVM implementation bugs too. (Google for JRockit and Hibernate issues for example).