It seems Cayenne requires one holistic model for an application (e.g. one
Cayenne.xml file). Is this true? Can I break up the model into multiple
sub-models (e.g. nodes and maps) and distribute throughout my class library?
For example, my architecture consists of multiple Eclipse projects (e.g.
foundation, various business logic domains, user interface, etc.), most of
which require their own model. These projects are subsequently combined as
required to form executables.
EOF had let me divide my model as required, and it would be dynamically
combine the sub-models into a larger holistic models as it located them
within the runtime. I would like to do something similar here.
So some questions:
1. Is there only one cayenne.xml file permitted per executable?
2. Can I hand-edit cayenne.xml to reference various nodes and maps in
various projects?
3. What will happen if cayenne.xml references 5 nodes in 5 distinct
projects, but only 3 of those projects are linked into an application?
On the latter point, I understand the need to ensure cross-model
relationships do not introduce "class not found" errors. But any other
concerns?
Thanks,
Randy
