When the new Model Driven MDG generation mechanism came in a long while back, we investigated some of the specialisation features - allowing stereotypes to inherit properties, etc., from others. We found it wasn't as orthogonal as we'd hoped, but it was serviceable enough to use what worked to make our MDG management lives a LOT easier. This mechanism has moved on to some extent, and it's unclear what the current "State of play" might be.
We've now got a situation where we have some similar relationships (basically, the same "functionality" but with different parameters) that we'd like to have a common rendering with additional widgets to indicate the parametric differences. For example,
- This item has a name that is identical to the name of that item
- This item has an alias that is identical to the alias of that item
- This item has a name that is identical to the alias of that item
- This item has an alias that is identical to the name of that item
Unfortunately, EA provides VERY limited mechanisms for managing this via the properties available to the relationship shapescript, and we have determined that the best solution in the current circumstances is to use separate stereotypes to identify the appropriate case.
In this example, the stereotypes might be:
- IdentityN2N
- IdentityA2A
- IdentityN2A
- IdentityA2N
(NOTE: All the relationships are reflexive, and the nominated directionality is from origin to destination)
So now you understand the context, we intend to have the same shapescript to generate the relationship arcs on the diagram, have the shapescript respond to the actual stereotype, and render the relationships slightly differently depending upon the specific relationship metatype involved.
My question is, can we
actually model it that way where the specific metatype has an inheritance relationship to a base metatype (with the shapescript defined there) and does NOT have a shapescript definition of its own, OR do we have to have a
simulation "kludge" where the same shapescript text is added to each of the specific metatypes (as we've had to do for other mechanisms).
TIA,
Paolo
