Subclassing in the model makes sense to me.
I don't think anything would need to be added to AbstractSCAObject -
by the time we get to the runtime all of the intents and policySets
should have been processed and converted into wires with the
necessary interceptors.
--
Jeremy
On Mar 9, 2007, at 7:26 AM, Mark I. Dinges wrote:
Another person and I are starting to add support for requires and
policySets attributes to the various elements in the model. One
possible direction we are considering is to have those model
objects that support requires and policySets attributes extend the
PolicyAttachableModel object. To this object a method would be
added to handle extracting the requires and policySets attributes
and would save them. Another option would be to have an instance of
a PolicyAttachableModel object in each of the model objects that
support the requires and policySets attributes. Does the community
have a preference for either of these options or have a suggestion
for another option? During the build phase the attributes would be
processed from the model. This is where the reduction and checking
for conflicts would take place. Methods would be added to
AbstractSCAObject to add/get intents and policsySets.This would
make it so that the various implementations of the model would
inherit them.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
