[OJB-developers] JUnit Tests & MySQL / global IDs
Brought to you by:
thma
From: <Joa...@tp...> - 2002-03-20 10:10:43
|
Hy, I've got several questions. First of all I still got the trouble with MySQL failing a Broker Test (PersistenceBrokerTest.testEscaping()) and it doesn't seem that this will be resolved, unless we change the Test. MySQL doesn't have a problem with the escaping but with trailing spaces (the get striped in the database). I suggest that the trailing spaces in the escaping Test are removed. The questions is whether we want to do the trailing-spaces-Test as well (which would again result in a Failure with MySQL, but at least it would be more precise). I'd be willing to make this Change, if no-one objects. Then there is my newly created Option for SequenceManagerHighLowImpl, which allows for OID that are unique over all classes (to be exact they are unique over all classes that use the same name for their ID-field. This was necessary to allow the JUnit tests to pass and will not be a problem, 'cause when you use this function you're likely to use an oid-field (or something) for all classes). The implementation cannot really handle Data being inserted into the DB without using it for acquiring the IDs with this function. So it is quite useless for accessing legacy data, but for our product it is just fine (we won't have any DB-access without OJB). The function is not yet configurable (except for setting globalSequence = true in the constructor and recompiling), as I don't to wait for the rewrite of BaseConfiguration (which I would volunteer to do, if no-one has started work on it yet, maybe if anyone has any design ideas. If not, I'll just do it how I see fit ;-) Well, basically most of my questions boil down to one point: are there any cvs-checkin rules that should be obeyed? Not breaking the JUnit tests seems to be an obvious test. Any others? How about new functionality, should it be discussed on this list before check in? Bug fixes? btw, anyone got a problem, when I check in a .cvsignore file into CVS that contains *.class? regards Joachim Sauer |