Software Product Line Engineering Basics - Model-Driven and Software Product Line Engineering - page 37

D EFINITION 2.6.- The variability of a set of software systems or
products is the set of differences, described in a structured way,
of some or all of their characteristics.
There are many methods for classifying and documenting
variability in software product lines [CHE 09]. Several
approaches for classifying and documenting variability focus
on the use of variability models [SIN 07, BAY 06].
From an abstract point of view, variability management
can be seen as the definition of variation points and variants .
A variation point is a location point in an artifact where
severalvariantsarepossible.Avariantisaconcreterealization
of a variation point; it corresponds to an alternative to
bind the variation point. There are some characteristics or
constraints related to the possible variants,which are attached
to the variation point. In a secure Smart-Home, we have
authentication requirements (the variation point),which could
be more or less sophisticated. The variants are a simple key
lock, an electronic keyboard, and various kinds of recognizer
devices (e.g. visual, vocal, digital). In the physical architecture
of the Smart-Home, we should allow various structures. Thus,
we can have one or more floors, each of them with different
structures in terms of rooms, doors, windows, etc. Another way
to describe this variability is to enumerate a set of predefined
alternatives such as standard flat, house, hotel, castle, etc. The
Smart-Home structure is a variation point with many complex
variants.
D EFINITION 2.7.- Variability management is the process of
making the variability of software artifacts explicit and to
enable variant binding throughout the lifecycle.
Pohl et al . [POH 05b] define variability management as
the set of activities for defining and exploiting variability
throughout the SPL development lifecycle. The concept
covers the following issues: (1) supporting activities
concerned with variability and commonality analysis, which