Bugzilla – Full Text Bug Listing
| Summary: | Property should not specialize TemplateableElement | ||
|---|---|---|---|
| Product: | [Modeling] MDT.UML2 | Reporter: | Rafael Chaves <eclipse> |
| Component: | Core | Assignee: | James Bruck <bruck.james> |
| Status: | CLOSED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | bruck.james, Kenn.Hussey |
| Version: | 2.2.0 | Keywords: | plan |
| Target Milestone: | M3 | ||
| Hardware: | PC | ||
| OS: | Windows Vista | ||
| Whiteboard: | Compliance | ||
|
Description
Rafael Chaves
You're right! This must have been missed when the generalizations were removed during UML 2.0 finalization. Unfortunately, we can't remove this generalization (and the associated properties) from the Property metaclass until the next major release (3.0); the best we can do in the meantime (i.e. in 2.2) is mark the corresponding methods and metadata deprecated and remove their property descriptors from the edit layer - I've just committed these changes. In theory, would it not be possible for the properties' type or the properties' qualifier to be items that could be involved in a template binding. Say you have 2 properties p and q. p has a template where its type is a parameter of kind Class. There could be a binding from q to the template of p that substitutes MyClass for the parameter. Although impractical, this could be one usage of a template on a property or am I way off base. Is the spec just negligent in mentioning this? I think that binding of the type for a property (like a parameter for an operation) would always occur within the context of an owning classifier (as with operations), meaning that the classifier should own the template signature... Committed 20081005. (Galileo M3) verified 20081104 |