Issues for UML 2 Diagram Interchange Revision Task Force

List of issues (green=resolved, yellow=pending Board vote, red=unresolved)

List options: All ; Open Issues only; or Closed Issues only

Issue 8971: Diagram Interchange clarification
Issue 10499: Section: Annex C
Issue 10660: Invalid "diagram-interchange.xml" file in DI 2.0 's ptc/05-06-07 zip file
Issue 13533: di-schema.xsd

Issue 8971: Diagram Interchange clarification (uml2di-rtf)

The following figure shall picture a Part of  a composite structure
diagram. Are the graphically informations like the colon that seperates the
name and the type of the part, the brackets that surround the multiplicity,
and the ".." points that display the range of the Multiplicity   legal
Diagram Interchange informations ? Till now I saved
those informations as TextElements with a SimpleSemanticModelElement.
Is this ok or can I disregard those informations ?


    ______________________________________
    |                                     |
    |     myPart : myPartType [1..*]      |                          
    |_____________________________________|

I was looking over the specification wanting to create a program that will be able to produce and read the format. However, I could use some clarification in Annex C, C.2.1 Class Diagram. There is no representation for an association. I had been, for a short time, believed that it was DiAssociationClass, but upon looking that up, I found it was not. For DiAssociationEnd, the tree does not match up with the example, as I understand it, in Annex B. Annex C says that the RoleAdornment should product a node with simpleSemanticModelElement of 'RoleAdornment'. But there is no such nesting. The GraphNode(<AssociationEnd>)'s next nestings go straight to 'name', 'visibility', and 'multiplicity'. Could you clarify this for me? Also in the zip example, there is an undefined compartment used called 'ExpressionCompartment'. DiAssociationClass refers to 'Stereotypes'. Is that supposed to be StereotypeCompartment? In C.1, there are a few areas where ',' appears that is should be '<'.

I tried to load the diagram-interchange.xml file in the referenced zip file into the Rational Rose UML 1.4 xmi addin.  This addin reports 4 errors: 
 
Could not add DataType "Image" to Logical View -- probably duplicate name.
Could not add DataType "Dimension" to Logical View -- probably duplicate name.
Could not add DataType "Point" to Logical View -- probably duplicate name.
Relation "Generalization" @  in 'Image' was not resolved -- not added.

 
The first three are cases where data type names duplicate classes in the same package.  This is at least bad form, if not actually illegal.   Rose does not allow it.
 
THe final error is a case where one of the "generalization" references of the "Image" class points to a Generalization object that is not in the file:
 
<UML:GeneralizableElement.generalization>

<UML:Generalization xmi.idref="EAID_B3BDA775_38C1_4989_87D1_A4CDF297DA06"/>

</UML:GeneralizableElement.generalization>

There is no object with an ID of "EAID_B3BDA775_38C1_4989_87D1_A4CDF297DA06" in the file.

Invalid "diagram-interchange.xml" file in Diagram Interchange 2.0 's ptc/05-06-07 zip file

The di-schema.xsd available from the link http://www.omg.org/cgi-bin/doc?ptc/05-06-07 import namespace <xsd:import namespace="http://www.omg.org/XMI" schemaLocation="XMI.xsd"/>. XMI.xsd is not present in the same package throws schema error on di-schema.xsd. Where can be the approved XMI.xsd found that the di-schema refers to.