1) what character encoding do I need to use for the import.sql file: ISO-8859-1 or UTF-8 2) when you have master-child data to insert, what is the best way to do it? since I am using MS-SQL database, the ID's will be generated by the database, so I must do this: insert into ParentTable (field1, field2, ..) values (..) | insert into Child Table (filed1, field2, .., ParentId) values (..., 1) | insert into Child Table (filed1, field2, .., ParentId) values (..., 1) | insert into ParentTable (field1, field2, ..) values (..) | insert into Child Table (filed1, field2, .., ParentId) values (..., 2) * this is not really nice, that I have to "guess" the parent id | * the other thing is, that when you ever change your database to e.g. oracle, you have to completely rewrite your importscript: e.g. in this case, you have to add the id field for every row | insert into ParentTable (id, field1, field2, ..) values (1, ..) | | insert into Child Table (id, filed1, field2, .., ParentId) values (1, ..., 1) | | insert into Child Table (id, filed1, field2, .., ParentId) values (2, ..., 1) is there a better way? maybe I should not use the updatescript at all and create the data on application startup in java-code - then I could insert my entities and didn't have to care about the underlying database, right? 3) is this right: the import.sql not really has anything todo with hibernate - it's just an importscript and hibernate automatically executes it against your database at startup (if configured) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4131069#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...