Software Development Process Modeling. Developers Perspective to Contemporary Modeling Techniques
Mäkilä, Tuomas (2012-11-02)
Software Development Process Modeling. Developers Perspective to Contemporary Modeling Techniques
Mäkilä, Tuomas
(02.11.2012)
Turku Centre for Computer Science
Julkaisun pysyvä osoite on:
https://urn.fi/URN:ISBN:978-952-12-2790-5
https://urn.fi/URN:ISBN:978-952-12-2790-5
Kuvaus
Siirretty Doriasta
Tiivistelmä
Formal software development processes and well-defined development methodologies
are nowadays seen as the definite way to produce high-quality software within
time-limits and budgets. The variety of such high-level methodologies is huge
ranging from rigorous process frameworks like CMMI and RUP to more lightweight
agile methodologies. The need for managing this variety and the fact that
practically every software development organization has its own unique set of development
processes and methods have created a profession of software process
engineers. Different kinds of informal and formal software process modeling languages
are essential tools for process engineers. These are used to define processes
in a way which allows easy management of processes, for example process dissemination,
process tailoring and process enactment.
The process modeling languages are usually used as a tool for process engineering
where the main focus is on the processes themselves. This dissertation
has a different emphasis. The dissertation analyses modern software development
process modeling from the software developers’ point of view. The goal of the
dissertation is to investigate whether the software process modeling and the software
process models aid software developers in their day-to-day work and what are
the main mechanisms for this. The focus of the work is on the Software Process
Engineering Metamodel (SPEM) framework which is currently one of the most
influential process modeling notations in software engineering.
The research theme is elaborated through six scientific articles which represent
the dissertation research done with process modeling during an approximately
five year period. The research follows the classical engineering research discipline
where the current situation is analyzed, a potentially better solution is developed
and finally its implications are analyzed. The research applies a variety of different
research techniques ranging from literature surveys to qualitative studies done
amongst software practitioners.
The key finding of the dissertation is that software process modeling notations
and techniques are usually developed in process engineering terms. As a consequence
the connection between the process models and actual development work
is loose. In addition, the modeling standards like SPEM are partially incomplete
when it comes to pragmatic process modeling needs, like light-weight modeling
and combining pre-defined process components. This leads to a situation, where
the full potential of process modeling techniques for aiding the daily development
activities can not be achieved.
Despite these difficulties the dissertation shows that it is possible to use modeling
standards like SPEM to aid software developers in their work. The dissertation
presents a light-weight modeling technique, which software development teams
can use to quickly analyze their work practices in a more objective manner. The
dissertation also shows how process modeling can be used to more easily compare
different software development situations and to analyze their differences in a
systematic way. Models also help to share this knowledge with others.
A qualitative study done amongst Finnish software practitioners verifies the
conclusions of other studies in the dissertation. Although processes and development
methodologies are seen as an essential part of software development, the
process modeling techniques are rarely used during the daily development work.
However, the potential of these techniques intrigues the practitioners.
As a conclusion the dissertation shows that process modeling techniques, most
commonly used as tools for process engineers, can also be used as tools for organizing
the daily software development work. This work presents theoretical solutions
for bringing the process modeling closer to the ground-level software development
activities. These theories are proven feasible by presenting several case studies
where the modeling techniques are used e.g. to find differences in the work methods
of the members of a software team and to share the process knowledge to a
wider audience.
are nowadays seen as the definite way to produce high-quality software within
time-limits and budgets. The variety of such high-level methodologies is huge
ranging from rigorous process frameworks like CMMI and RUP to more lightweight
agile methodologies. The need for managing this variety and the fact that
practically every software development organization has its own unique set of development
processes and methods have created a profession of software process
engineers. Different kinds of informal and formal software process modeling languages
are essential tools for process engineers. These are used to define processes
in a way which allows easy management of processes, for example process dissemination,
process tailoring and process enactment.
The process modeling languages are usually used as a tool for process engineering
where the main focus is on the processes themselves. This dissertation
has a different emphasis. The dissertation analyses modern software development
process modeling from the software developers’ point of view. The goal of the
dissertation is to investigate whether the software process modeling and the software
process models aid software developers in their day-to-day work and what are
the main mechanisms for this. The focus of the work is on the Software Process
Engineering Metamodel (SPEM) framework which is currently one of the most
influential process modeling notations in software engineering.
The research theme is elaborated through six scientific articles which represent
the dissertation research done with process modeling during an approximately
five year period. The research follows the classical engineering research discipline
where the current situation is analyzed, a potentially better solution is developed
and finally its implications are analyzed. The research applies a variety of different
research techniques ranging from literature surveys to qualitative studies done
amongst software practitioners.
The key finding of the dissertation is that software process modeling notations
and techniques are usually developed in process engineering terms. As a consequence
the connection between the process models and actual development work
is loose. In addition, the modeling standards like SPEM are partially incomplete
when it comes to pragmatic process modeling needs, like light-weight modeling
and combining pre-defined process components. This leads to a situation, where
the full potential of process modeling techniques for aiding the daily development
activities can not be achieved.
Despite these difficulties the dissertation shows that it is possible to use modeling
standards like SPEM to aid software developers in their work. The dissertation
presents a light-weight modeling technique, which software development teams
can use to quickly analyze their work practices in a more objective manner. The
dissertation also shows how process modeling can be used to more easily compare
different software development situations and to analyze their differences in a
systematic way. Models also help to share this knowledge with others.
A qualitative study done amongst Finnish software practitioners verifies the
conclusions of other studies in the dissertation. Although processes and development
methodologies are seen as an essential part of software development, the
process modeling techniques are rarely used during the daily development work.
However, the potential of these techniques intrigues the practitioners.
As a conclusion the dissertation shows that process modeling techniques, most
commonly used as tools for process engineers, can also be used as tools for organizing
the daily software development work. This work presents theoretical solutions
for bringing the process modeling closer to the ground-level software development
activities. These theories are proven feasible by presenting several case studies
where the modeling techniques are used e.g. to find differences in the work methods
of the members of a software team and to share the process knowledge to a
wider audience.
Kokoelmat
- Väitöskirjat [2926]