|
|
Re: UML Testing profile [message #472680 is a reply to message #472654] |
Thu, 19 April 2007 21:06 |
Eclipse User |
|
|
|
Originally posted by: carrasco.ModelDrivenDevelopment.co.uk
Here we are again.
That is not the standard view of the metamodeling stack.
That is the viewpoint of those that have invested heavily in UML,
and want to avoid dispersion into many of what we now know as "Domain
Specific Languages"
(like testing).
The other viewpoint, of those whose drive is not aligned with UML industry,
is that the concepts are directly captured as instances of MOF,
but, if you have a UML tool,
you may use the lightweight extension mechanism
called "Profiles" (stereotypes onto UML classes).
Many use to do both: the metamodel as instance of MOF
(for conceptual undestanding, and driving model-driven-engines and
generators like EMF, GMF)
then also a UML Profile, for those that use Rational Rose, and the like.
That
|
|
|
|
Re: UML Testing profile [message #608519 is a reply to message #472654] |
Thu, 19 April 2007 21:06 |
Eclipse User |
|
|
|
Originally posted by: carrasco.ModelDrivenDevelopment.co.uk
Here we are again.
That is not the standard view of the metamodeling stack.
That is the viewpoint of those that have invested heavily in UML,
and want to avoid dispersion into many of what we now know as "Domain
Specific Languages"
(like testing).
The other viewpoint, of those whose drive is not aligned with UML industry,
is that the concepts are directly captured as instances of MOF,
but, if you have a UML tool,
you may use the lightweight extension mechanism
called "Profiles" (stereotypes onto UML classes).
Many use to do both: the metamodel as instance of MOF
(for conceptual undestanding, and driving model-driven-engines and
generators like EMF, GMF)
then also a UML Profile, for those that use Rational Rose, and the like.
That
|
|
|
Powered by
FUDForum. Page generated in 0.04124 seconds