Issue 3123: "Physical" Metamodel Package Structure (uml-rtf) (uml2-superstructure-ftf)
Source: Escape Velocity (Mr. Don Baisley, donbaisley(at)live.com)
Nature: Uncategorized Issue
Severity: 
Summary: The package structure of UML 1.3 makes it difficult to deploy small parts of
the "physical" metamodel separately.  For example, a MOF-based facility that
supports classes from Behavioral_Elements.Common_Behavior must support all
of Behavioral_Elements.  A facility that supports Exceptions must also
support Use Cases and State Machines.  This has been a problem in the
formation of the CWM metamodel which extends UML.  Its interfaces and DTDs
are made to be much too large.

The result of UML currently having three metamodels (two of which are large)
rather than many smaller metamodels is that the IDL modules are very large
and so are the DTDs.  Breaking the metamodels into several smaller ones will
allow smaller interface sets and DTDs that can be mixed and matched to
provide necessary functionality without a huge overhead.  

Resolution: 
Revised Text: 
Actions taken:
December 15, 1999: received issue
March 9, 2005: closed issue
Discussion: 
The package structure of the UML 2.0 is radically different from UML 1.x, so that
this issue is no longer applicable.
End of Annotations:=====
From: "Baisley, Donald E" <Donald.Baisley@unisys.com>
To: issues@omg.org
Subject: "Physical" Metamodel Package Structure (uml-rtf)
Date: Wed, 15 Dec 1999 11:57:25 -0600
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: text/plain;
	charset="iso-8859-1"
X-UIDL: eY~e9;k)!!&14!!1?;!!

The package structure of UML 1.3 makes it difficult to deploy small
parts of
the "physical" metamodel separately.  For example, a MOF-based
facility that
supports classes from Behavioral_Elements.Common_Behavior must support
all
of Behavioral_Elements.  A facility that supports Exceptions must also
support Use Cases and State Machines.  This has been a problem in the
formation of the CWM metamodel which extends UML.  Its interfaces and
DTDs
are made to be much too large.

The result of UML currently having three metamodels (two of which are
large)
rather than many smaller metamodels is that the IDL modules are very
large
and so are the DTDs.  Breaking the metamodels into several smaller
ones will
allow smaller interface sets and DTDs that can be mixed and matched to
provide necessary functionality without a huge overhead.  

Recommendation:  Make each nested package in the UML metamodels into
its own
metamodel.  But preserve the current structure in the hierarchical
identification of the packages.  For example, Common_Behavior is
currently a
nested packaged within the metamodel that is fully identified as
org.omg.uml.Behavioral_Elements.  I recommend we make Common_Behavior
(and
each of its brothers) into a separate MOF metamodel identified as
org.omg.uml.behavioral_elements.Common_Behavior.

Breaking the large composite metamodels into smaller ones will solve a
problem encountered in modeling CWM, which uses only two classes from
all of
Behavioral_Elements.


Don Baisley
Unisys