Hardware Reference
In-Depth Information
An attractive way to document proven solutions to specific development process
problems is to use organizational patterns [5] (or process patterns [6]). A collection of
such solutions can be further organized into process pattern languages [7]. A process
pattern language need not cover the entire process, but it can concentrate on a certain
viewpoint of the software development process. In this work the viewpoint is GSD:
we derive a pattern language for project management in GSD (GSD Patterns). The
solutions in these patterns have been mined from the practices that have been found to
work well in a large company operating in the field of process automation.
In general, patterns represent knowledge that is validated by previous experience.
However, if patterns are mined from a limited environment, as in our case, this argu-
ment does not hold. In this work we have evaluated the resulting patterns by using a
scenario-based technique introduced in [8].
This paper is organised as follows: The next section describes our research ap-
proach, and the methods used. Section 3 presents the GSD pattern language. Section 4
discusses the evaluation results of the GSD pattern language. Finally, we discuss
related work in Section 5 and conclude the paper in Section 6.
2 Research Approach
This section introduces the research approach which includes both the collection of
process patterns and an assessment method for the pattern language.
2.1 Collecting Process Patterns
GSD patterns presented in this paper have been collected from industry and literature
during studies [9, 10, and 11] and the pattern evaluation meetings in a large company
operating in the sector of the process automation industry. The pattern evaluation
meetings have also been organised with other companies. The total number of re-
spondents in these studies has been 32 in questionnaires and 25 in interviews.
Each separate study has been started by choosing a certain software process area in
which problems and best practices have been collected from the viewpoint of global
software development. The best practices have been presented in the form of GSD
patterns. In each study, the collection of case data has been done by using question-
naires and interviews. The framework for data collection is organised based on the
concepts, practices or phases of the development process (referred to as framework
items in the sequel) depending on the study area. For each framework item there have
been three open-formed questions: what is good, what needs improvement and how to
improve if there were no restrictions in the implementation. After the questionnaire,
key persons were interviewed to get more detailed information about the case. The
persons selected for the interviews represented project managers, product managers
and project members. The interviewees worked in the company or in its partner com-
panies. The framework was used as a checklist for the interviewer, leaving room for
open discussion. Questionnaires and interviews produced raw data for analysis. The
raw data has been processed and analysed by organising it based on the framework
items. After that, proposed process patterns were created based on processed case
Search WWH ::




Custom Search