Information Technology Reference
In-Depth Information
complextogetstructural,functional,anddynamicdescriptions
together. Generally, models provide structural aspects; some
behavioral constraints can be defined using OCL. However,
even when OCL facilitates the description of semantics of
methods,itdoesnotmakethegenerationofefficientcodeeasier.
Some approaches are able to cope with dynamic behavior but
consistency of the overall descriptions and code generation
remain mainly under study.
Many of the benefits expected from software product lines
are based on the assumption that the additional investment in
setting up a product line pays off later when products are built.
Thus,to define the basis for model-driven and software product
line engineering, one must integrate both approaches to define
new processes in the
problem space
for expressing variability
and configuring products,and in the
solution space
for deriving
products. MD-SPL deeply impacts the management of core
assets,theexpressionanduseofvariability,andtheproduction
chain.
Regarding the problem space, the issues are “How to define
mechanisms to manage variability based on metamodeling;
for e.g. feature modeling? How to implement multi-staged
configuration of products? How to model and configure
products from several concerns and several abstraction
levels?”
An effective return on investment in product line
engineering is achievable when the product lines can be
efficiently used for product derivation. If we want to improve
product derivation, we require models that are more than just
vehicles for documentation and discussions on the whiteboard.
Then, regarding the mechanisms used in the solution space,
questions that arise are: What are the core assets that
are required? How to create and use decision models in
conjunction with other model assets to tackle the derivation
of products?
Search WWH ::
Custom Search