> Only that I don't see why AuthorId is now being generated as an int rather > than a NumberKey. Exactly what was done to cause this? As far as I can > tell the test case was correct when it used NumberKey - the failure of a > test case doesn't necessarily mean the test case is wrong after all.
There was a big change put through this weekend where it used to be that getters/setters for primary keys (e.g. getAuthorId) would have to take some variant of ObjectKey (I think that's the base key class). Now they take their native types. The author id in the database is an integer, so it was decided that it makes more sense for getAuthorId/setAuthorId to also take an integer, just like the other getters/setters. Hence the LargeSelectTestCase had to be changed. Note that getPrimaryKey will still return the ObjectKey version of primary keys. - Stephen -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
