I've allocated this bug report to the tool but the choice is arbitrary. It can also be assigned to the methodology. The bug refers to a contradiction between the rules for schema mapping of extensions as described in section 4.6 of the methodology document and what the tool does in reality.

Section 4.6 states:

Classes which are connected to other classes via an aggregation/composition using a UML qualifier have by default multiplicity of 0..* ("Note that in this particular case ―LowerBound‖ is hardcoded to '0'"). Hence, they are optional but there is noe way to deselct such classes in the tool, i.e. they are handled by the tool like mandatory classes.

User tori encountered a consistency problem with the usage of the term "subscription" between the Exchange PSM and the Developers guide. He raise the issue as a question in the forum, but since it is an obvious bug in the documentation, I re-enter it as an issue:

"Wed, 04/08/2010 - 10:40 | tori
 
Hello,

If you have two extensions that extends the same class (like CCTV extends general publication), and then in selection feature selects to just generate one of them will not be reflected in the schema. If you do the _XXExtensionType will have all "children" independent if they were selected or not. This leads to that a not valid schema is created.

Solution is to add a additional check when generating the _XXXExtensionType, if class should be generated or not.

If you in the model entes extension name and extension versions those attributes are generated in the output schema.

The following packages have no tag values and hence no definitions:
PointAlongLinearElement
Humidity
Pollution
PrecipitationDetail
RoadSurfaceConditionMeasurements
Temperature
Visibility
Wind

This was not detected by the schema generation tool. But also there does not seem to be any statements in the methodology about mandatory tags for packages other than certain special packages. Should this not be considered for completeness.

Hence this may impact the tool and methodology.

Hi,
I've tried to add a small LevelB extension (a new value for the MeasuredDataPublication) to the model. The D2Convertion tool reads the resulting xmi-file without complaints but if I try to change to the selection tab it throws an unhandled exception.
------
See the end of this message for details on invoking
just-in-time (JIT) debugging instead of this dialog box.

When clicking right on the selection tree, you can choose a filename to save the current selection. Immediatly after that, the program gives an unhandled exception (see attached image).
It's possible to continue without loosing data, but no saving is done.

I'm using Windows 7.

Thank you!
Jörg Freudenstein

Here comes the debug-message:

See the end of this message for details on invoking
just-in-time (JIT) debugging instead of this dialog box.

************** Exception Text **************
System.NullReferenceException: Object reference not set to an instance of an object.

In the selection tree, it's not possible to disselect two entrys at the same time. Whenever you disselect one of them, the other one becomes selected automatically and vice versa, although there shouldn't be any correlation.

I can reproduce this bug with the following two entries (in Situation - SituationRecord), but I think there are more faulty combinations like this):

1) Validity - OverallPeriod - validPeriod
2) GroupOfLocations - Location - NetworkLocation - destination

Thank You!
Jörg Freudenstein

The Specific datatype "AngleInDegrees" should use the facets tag to restrict its valid value range to be 0 - 359 inclusive making it consistent with its definition.