If a folder with documents is copied then the onCopyComplete-Policy is not called in the exact order of the childs.
In our project we have a domain specific indentifier for children of a folder, which is added via a behaviour.
If a folder is copied, then this identifier is the same as in the copy-source.
The children are copied in the right order, but the identifier is added randomly.
I debugged this and found out that the problem is the implementation of an map in the CopyServiceImpl.copy method. This method holds all copied nodes on which later the onCopyComplete-Policy is executed. But this Map is a HashMap, which order is defined by the hashValue of its entries.
I thought this might be the best way to contribute my first code-changes via a pull-request.