Hi Ed,
Thank you very much for this wiki entry. It's very helpful and it
brilliantly exposes the great work you and Kenn have been doing around
the integration of OCL inside Ecore models.
Here is my feedback:
- I have done minor changes/corrections in the wiki entry.
- I would avoid mentioning "Ecore-Constraints" and "Ecore-Invariants"
terms, since it may be confusing and it doesn't fit from an OCL point
of view. If EMF provides a different way to validate an EClassifier
instances, it is not definetely part of a guide about how to integrate
OCL constraints in an Ecore model, specially when this way
("Ecore-Invariants") doesn't provide any additional enhancement. It
only provokes the user to think about the the terms and the difference
between an Ecore-constraint and an Ecore-invariant. Furthermore, an OCL
invariant is thought to be applied to an EClassifier, while an
"Ecore-Invariant" could only be applied to EClasses. Therefore, I
would only mention the term EClassifier Invariants and I'd suppress the
"Ecore-invariant" case.
- If you provided in a branch or somewhere the OclInEcore editor, I
could try to install the example and fullfill the missing
<Description to be provided.> part.
Apart from that:
1. I think that the old-fashioned documentation article which explained
how to introduce OCL constraints in a metamodel using custom JET
templates should be replaced by this one when it's completed. I could
invest some time to do that.
2. I think there were some open bugzillas recalling an update of the
documentation article mentioned above. I'll find them to solve.
Cheers,
Adolfo.
|