Issue 6456: Use 'represent' for the relationship of a model (uml2-rtf)
Source:  (, )
Nature: Uncategorized Issue
Severity: 
Summary: "An instance specification is a model element that represents an instance in a modeled system." [7.7.1]  That is, the relationship of the instanceSpecification with a class to an object of that class is the representation relationship.


At the same time, a lifeline represents a connectable element. [14.2]


This is an example of a recurrent problem in the specification: model elements that represent other model elements.


At the same time, "attributes of a class are represented by instances of Propert[ies]..."


This is an example of an occassional and quite striking problem in the specification: items in the modeled system that represent model elements.


The theory of representation needs to be settled.  That done, the specification needs to be reviewed with this in mind and all improper uses of representation corrected.


Resolution: 
Revised Text: 
Actions taken:
November 7, 2003: received issue
August 23, 2006: closed issue
Discussion: 
Indeed. It would be useful if the theory of representation was settled. However, this is a
deep issue that exceeds the scope of an F/RTF. Hence, it is better to defer this issue to a
subsequent revision of UML. The submitter is absolutely right and that more care should have been taken in the use of
this term and many others. Unfortunately, the problem is endemic in both the Infra and
Super specs (and, likely, many other OMG specs). A count reveals close to 500 uses of
the word “represent” or its derivatives scattered throughout the document. Each of these
would have to be examined and appropriate action taken. Of course, this would only
solve this particular case of loose terminology but not the overall problem. The FTF
punted this issue by deferring it. However, it is unrealistic to expect that any future RTF
will have the resources or the time to fix this particular problem, let alone the more
general problem.
What is really required is a set of OMG-wide style guidelines and, possibly, a glossary of
approved defined terms that should be used in writing all submissions related to
modeling. This is a problem that exceeds this particular specification and should be posed
as an issue to the OMG as a whole.
Furthermore, although the loose use of “represents” is confusing in certain spots, in most
cases, it is not a problem that will lead to errors in tool implementation or significant
misinterpretation by readers. Therefore, it is proposed to close this issue in the context of
this specification. Disposition: Closed, no change
End of Annotations:=====
nstance specification is a model element that represents an instance in a modeled system." [7.7.1]  That is, the relationship of the instanceSpecification with a class to an object of that class is the representation relationship.


At the same time, a lifeline represents a connectable element. [14.2]


This is an example of a recurrent problem in the specification: model elements that represent other model elements.


At the same time, "attributes of a class are represented by instances of Propert[ies]..."


This is an example of an occassional and quite striking problem in the specification: items in the modeled system that represent model elements.


The theory of representation needs to be settled.  That done, the specification needs to be reviewed with this in mind and all improper uses of representation corrected.


...


Suggested resolution:  Use 'represent' for the relationship of a model to that which the model represents (for example, the often mentioned "modeled system").