Well, I've discovered another one (at least in 1.2, which implies 2.0,
also), but haven't tracked it down enough yet. I don't think it has
anything to do with our binary PKs, either. In a nutshell, I had optimistic
locking turned on everything, including relationships and had:
A ->> B (A to-many-to B)
When I updated attributes in A and B and committed the changes, I'd get an
optimistic locking exception because the SQL generated looked like:
UPDATE A set foo = 'bar'
WHERE primaryKey IS NULL AND ...
For some reason, it is nulling out the PK when figuring out the
differences. (Strangely, it doesn't null out the PK in B -- it was updating
before the A record and the PK was there.) I just haven't had time to track
it down. My solution was to uncheck optimistic locking.
I'm not opposed to a new release, but I know there is something lurking
there, too.
Thanks,
/dev/mrg
On 4/27/07, Andrus Adamchik wrote:
I wish our next release could have been 3.0M1, and we got very close.
For my part I need about two weeks of uninterrupted work on Cayenne -
something I can't afford now :-/
Nevertheless, we have 13 closed bugs on each of the stable branches
(1.2 and 2.0), so I think this is a good time to synchronously
release 1.2.3 and 2.0.3. Unless we hear from committers who are still
working on some stable issues (or some other objections), I'll start
a vote thread shortly.
Thanks
Andrus