Model-Driven Engineering - Model-Driven and Software Product Line Engineering - page 96

The first is about the representation of configuration models;
it must cope with the complexity of variability approaches.
Feature model configuration is rather simple; however, it is
moredelicatewithmetamodels,whichallowmorepowerfuland
constructivemeans.Theseconduseisinthederivationprocess;
it should consider the core asset and the configuration model
to generate the product. This mechanism should allow the
selectionoftransformationrulesandcontroloftheirscheduling
since there could be some feature or variation interactions.
The complexity of the software process makes the use
of traceability to track and analyze the flow of information
mandatory.Onechallengeisthemanagementofahugeamount
ofcomplexdata.Todoapreciseanalysis,weneedrelevantdata,
which means producing and processing this data.We also need
a tool supporting most of the stages in a product line, that is,
covering end-to-end traceability from market requirements to
test cases.
3.8. Summary
The model-driven engineering paradigm organizes the
whole software development cycle as a process of creation,
iterative refinement, and integration of models. Models are
first-class entities that capture a partial view of a system. Each
model conforms to its metamodel, which describes the domain
grammar and constraints. The unified modeling language is
often used to describe the structure of models and metamodels.
The object constraint language is the standard language to
express constraints over models and metamodels. It is a side
effect free and object-oriented language style devoted to the
description of UML elements and navigation in diagrams.
Domain-specific modeling is a way to develop software
systems that involves the use of domain-specific modeling
languages to represent the different concerns of an application
domain. In the context of MDE and DSML, we introduce the
concept of metamodels, the relation of conformance between