Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

AIB 2015-02: Testing Life Cycle-related Properties of Mobile Applications

1 view
Skip to first unread message

Thomas Ströder

unread,
May 10, 2015, 6:28:12 PM5/10/15
to
The following technical report is available from
http://aib.informatik.rwth-aachen.de:

Testing Life Cycle-related Properties of Mobile Applications
Dominik Franke
AIB 2015-02

With an increasing number of mobile devices like smartphones and
tablets, their relevance to users and growing number of available
applications, also their field of application widens. For the software
quality of mobile applications, the application life cycle - the
process-related states and state transitions - plays an important role.
Today's mobile platforms, like Android, iOS and Windows Phone, have
specific scheduling policies on application level to ensure the
reactiveness of an application, targeting an improved responsiveness and
a good user experience. Depending on the life cycle state of an
application, it is allowed or restricted to access resources like RAM
and CPU. Such policies can lead to data loss and unexpected behavior of
the mobile application.

This work presents a conceptual approach for testing application
properties which are related to life cycle state changes, so called life
cycle-related properties. The first step consists of reverse engineering
the life cycles of mobile applications. These life cycles are used as a
basis for testing life cycle-related properties at state changes. The
testing approach uses callback-mechanisms of the underlying mobile
platforms to check assertions about life cycle-related properties. It
handles application components with an own life cycle as units and tests
each unit in a unit-based testing approach. In a case study, the
conceptual approach is implemented for the mobile platform Android. One
of the results of the case study is the AndroLIFT tool for testing life
cycle-related properties of Android applications.

The evaluation of this work presents the capabilities and limitations of
the conceptual approach. While the approach is well-suited for today's
mobile platforms, extensible and scalable with respect to the type and
number of life cycle-properties, it mainly depends on the
callback-mechanism of the underlying mobile platform. The evaluation of
the AndroLIFT tool in the context of a practical course with student
participants confirms the value of the Android implementation of the
presented approach to test life cycle-related properties of Android
applications.

0 new messages