Here's the thing about "backwards compatibility"... That's only true if
the behavior is documented and/or tested as such or if the behavior is just
"inherently reasonable".
As you even seem agree, the previous behavior is not inherently reasonable;
its at best questionable. So then can you point
Ok. I will move this discussion to the users group.
I understand that it sounds reasonable to have a null check but the
important point is that it has broken backwards compatibility. Any thoughts
on that?
On Thu, May 8, 2014 at 7:07 PM, Steve Ebersole wrote:
> This list is for discussions about
This list is for discussions about the development of Hibernate, not for
usage discussions.
The behavior you describe sounds the most reasonable to me actually, tbh.
Also, generic code can (should, I'd argue) still do null checks...
On Thu, May 8, 2014 at 6:51 AM, amit shah wrote:
> We upgrad
