AbstractNew Advances in mobile computer technology and

the rapid growth of wireless networks in quality and quantity has

introduced new applications and concerns in computer science
and industry. The unique requirements and constraints
associated with mobile systems have brought new challenges to
software development for such environments, as it demands
extensive improvements to traditional systems development
methodologies in order to fulfill the special needs of this field.
We examine the challenges of developing software for mobile
systems, starting by reviewing mobile systems characteristics
and investigating the status quo of mobile software development
methods. It has been shown that Agile methodologies are
appropriate methods for the development of such systems; based
on this assumption, we identify specific requirements for a mobile
software development methodology, based on which a new agile
method is engineered using the Hybrid Methodology Design
approach. We claim that this methodology, and the approach
used for its construction, can facilitate the application of a
software engineering approach to the production of mobile
software systems.

Index Terms Mobile Software Development, Method
Engineering, Agile Methods

I. INTRODUCTION
EW advances in mobile computer technology are too fast
in emergence for the software engineering field to keep
up with. The number of mobile devices with computation
capabilities incorporated (such as third-generation mobile
phones and personal digital assistants) is growing all over the
world, and the ever-increasing demand for specialized
software for these devices has caused new concerns for
software developers, as this type of software has its own
unique characteristics and requirements.
Although commercial mobile systems have not been as
successful as originally predicted, mobile operators and
mobile value-added software providers expect that the
deployment of 3G technologies would have a dramatic effect
on the mobile applications industry, by expanding the vast

Manuscript received December 27, 2007. This work was supported in part
by the Research Vice-Presidency of Sharif University of Technology.
Vahid Rahimian is with the Computer Engineering Department, Sharif
University of Technology, Tehran, Iran (e-mail: [email protected]).
Raman Ramsin is an Assistant Professor at the Computer Engineering
Department, Sharif University of Technology, Tehran, Iran (e-mail:
[email protected]).

usage of mobile commerce applications and services.
Examples of such commercial applications include: mobile
information services, advertisement, content-based services,
location-based services, and mobile payment applications.
Development of a mobile software system differs from
traditional software development in many aspects, as mobile
software should satisfy special requirements and constraints.
Along with these specific constraints, software produced for
mobile environments should be at a high level of quality, so
that it can operate properly on different mobile devices that
exist or are expected to hit the market in the future. There are
numerous challenges that the designer of a software system for
mobile environments has to cope with. These challenges
mainly stem from ([1], [2]):
Wireless communication issues (considerations such as
availability and disconnection, bandwidth variability,
heterogeneous networks, and security risks);
Mobility issues (concerns such as address migration, and
management of location-dependent information);
Portability issues;
Various standards, protocols and network technologies;
Limited capabilities of terminal devices (factors
pertaining to low power, risks to data integrity, small-
sized user interfaces, and low storage capacities);
Special privacy and customizability needs;
Strict time-to-market requirements.
Some of these issues are due to deficiencies in current
technology, but most of them are intrinsic to mobility. The
design of mobile software systems is therefore much more
complicated than that usually seen in software development
projects, thus forcing developers to reconsider the use of
traditional software development methodologies.
Despite the above-mentioned problems, research endeavors
aimed at ameliorating the status quo through
enhancing/devising methodologies for mobile-software
systems development have been relatively few and far in
between. Most of the work performed in this field has been
focused on low-level (implementation-oriented) aspects of
software development, while high-level (methodology-
oriented) issues still remain to be properly addressed.
Designing an Agile Methodology for Mobile
Software Development: A Hybrid Method
Engineering Approach
Vahid Rahimian, Raman Ramsin
N
We propose a new agile methodology developed through
applying a Method Engineering (ME) approach [3]. The
approach applied herein, called Hybrid Methodology Design
[4], is used for iterative-incremental development of
methodologies based on a predefined set of requirements and
the knowledge acquired from existing methodologies and
process patterns/metamodels. We have elicited the high-level
requirements of a mobile-software development environment
through examining the available literature, and have fed it into
the Hybrid process to produce the target methodology. Ideas
from the ASD (Adaptive Software Development)
methodology [5] and New Product Development (NPD) [6]
were used in constructing the methodology.
The rest of the paper is structured as follows: section 2
reviews the related research so far performed on mobile
software development; in section 3, the appropriateness of
agile methods in software development for mobile
environments is discussed; in section 4, we identify the
requirements of a methodology tailored for mobile software
development; in section 5, the Hybrid Methodology Design
approach is used for building a new agile methodology for
developing mobile software, using the general patterns
followed by agile software development processes, and ideas
from ASD and NPD. The final section of the paper presents
the conclusions and opportunities for furthering this research.
II. RELATED RESEARCH
Identifying mobile computing challenges [7], [8],
adaptation of mobile software [9], [10], the notion of mobile
agents [11], resource sharing in mobile environments [12],
abstract architectures with special quality features [13], and
design patterns for developing mobile software [14] are
examples of the more popular, lower-level research so far
performed on mobile software development, whereas research
on specialized development methodologies is fairly limited. In
the rest of this section, we will provide a general survey of
methodology-level research conducted in the context of
mobile software development.
Gerstheimer and Lupp [15] take the end-user perspective to
propose a need-driven system design method for developing
mobile applications. Vainio et al. [16] recognize the use of
market elements as a leading factor in the success of a mobile
software product; they recommend utilizing market-based
New Product Development (NPD) for improving current
methodologies for use in mobile software development.
New Product Development is a complete process for
bringing a new product or service to the market [6]. There are
two parallel paths followed in the NPD process: one involves
idea generation, product design, and detailed engineering; the
other involves market research and analysis.
Vainio et al. have also analyzed the activities performed in
two commonly-used methodologies and have compared them
with two NPD process models. They have concluded that
current process models should contain more market-oriented
activities in order to be efficiently incorporated into the
development of mobile software products [16].
Another major work in the field of mobile software
development is that conducted by Abrahamsson et al. [17], in
which a methodology called Mobile-D is proposed as an agile
approach to mobile application development. The approach is
based on development practices borrowed from XP (eXtreme
Programming), enjoys method scalability inspired by the
Crystal family of methodologies, and provides life-cycle
coverage as prescribed by RUP (Rational Unified Process)
[17]. Through its phases and disciplines, Mobile-D tries to
merge the classic software development process (i.e.
traditional plan, design, implement, test, and release activities,
as mapped to Mobile-D disciplines of Phasing, Architecture
Line, Test-Driven Development, Continuous Integration, Pair
Programming, and Off-Site Customer) with the necessary
umbrella management/support processes (i.e. project
management, software configuration management, and
software process improvement, as mapped to Mobile-D
disciplines of Metrics, Agile Software Process Improvement,
and User-Centered Focus). Although the work of
Abrahamsson et al. on mobile software development seems to
be very promising, the description that they provide of their
Mobile-D approach is cursory and incomplete.
III. AGILE DEVELOPMENT: APOTENTIAL SOLUTION FOR
MOBILE SOFTWARE DEVELOPMENT
Our investigation of the problem and related research
indicates that agile methods have a good level of suitability for
the development of mobile applications. Although many agile
methods have been introduced over the last decade, none of
them has focused on the special requirements of mobile
software development. Yet as discussed in the rest of this
section, agile methods possess certain properties that make
them applicable to the mobile software domain.
Boehm and Turner [18] recognize five main factors that
affect agility: operating culture, team size, criticality of the
software, competence of the developers, and stability of the
requirements. Boehm argues that a software development
method works best when it is applied to situations with
specific traits [19]; he calls these situations the home ground
of the software development method. Table I compares the
home grounds for agile and plan-driven methods.
Abrahamson identifies agile methods as a potential solution
for mobile software development [20]. Based on the home
ground for agile methods, he performs a comparative analysis
to prove the suitability of agile methods for development of
mobile software, the results of which are shown in Table II.
While it has been suggested that agile methods are the most
appropriate means for mobile software development among
current software development methods, the special
characteristics of mobile devices and mobile networks demand
some adjustments to current development methodologies. In
fact, in order to fulfill the special requirements of mobile
software development, new methodologies are needed.
In the next section, we elaborate on the requirements of an
ideal software development methodology for mobile software
development.
IV. CHARACTERISTICS OF AN IDEAL MOBILE SOFTWARE
DEVELOPMENT METHODOLOGY
In this section, we propose the traits that we believe a
development method should have in order to be efficiently
employed for mobile software development. Based on the
properties we identify herein, a new methodology is later
constructed, using the Hybrid Methodology Design approach.
A. Agility
As stated in the previous section, agile methods seem to be
a good starting point for constructing a mobile software
development method. Agile methodologies are believed to
enhance software development flexibility and productivity, by
providing means to adapt to changes in requirements and the
environment, and also to learn from development experiences.
To support early and quick delivery of working software,
these methodologies use iterative-incremental development
engines to produce artifacts tangible to the customer.
Agile characteristics of highest importance in the context of
mobile software development include: iterative and
incremental process (which leads to enhanced risk-
management capabilities), test-driven development, adaptive
process, continuous customer involvement, highly skilled
developers, enhanced quality assurance, and continuous
process-wide reviews. Furthermore, considering the
competitive market for mobile software, shorter time-to-
market is a precious advantage; this could be achieved via
early releases of operational software, which is itself an
important feature of agile methods. Prioritization of
requirements is another agile practice that can prove essential
in mobile software development, since it sets the stage for and
governs risk management activities, and helps ensure that
features of higher value to the customer take precedence.
B. Market Consciousness
As the current market for mobile software is biased towards
relatively fine-grained software products, a general mobile
development process should be mainly oriented towards
product development, rather than project development.
Consequently, such processes should focus on establishing the
business case, thereby striving to identify the potential market.
The use of NPD practices for market analysis can enhance
mobile-software development: NPD process activities utilize
market information for mitigating uncertainties and risks. In a
market-oriented process for mobile software development,
market and customer needs should be carefully analyzed, and
a strict release schedule, which meets time-to-market
requirements, should be established and maintained during
development. In contrast to the typical modus operandi, in
which a processs main focus is on technical activities, a
mobile software development process should maintain a
balance between market-oriented and technical activities.
C. Software Product Line Support
A software product line is a set of software intensive
systems sharing a common, managed set of features that
satisfy the specific needs of a particular market segment or
mission and that are developed from a common set of core
assets in a prescribed way [21]. The special benefits of
applying the product line approach to mobile software
development are mainly due to the fast pace of advances in
mobile technology. This makes the life-cycle of mobile
software products ever shorter. Consequently, software
companies tend to develop a family of mobile software
products in a bid to reduce development costs.
As a result, we suggest that a mobile software development
process should provide means for supporting product line
engineering. Component-based development, use of
reconfigurable architectures, and product scoping can help in
achieving this capability. The process should also provide
adequate guidance on performing quality product line design.
TABLE I
HOME GROUNDS FOR AGILE AND PLAN-DRIVEN METHODS (ADAPTED FROM
[19])
Area Agile Methods Plan-Driven Methods
Developers
Agile, knowledgeable,
collocated, and collaborative
Plan-oriented; adequate skills;
access to external knowledge
Customers
Dedicated, knowledgeable,
collocated, collaborative,
representative, and empowered
Access to knowledgeable,
collaborative, representative,
and empowered customers
Requirements Largely emergent; rapid change Knowable early; largely stable
Architecture
Designed for current
requirements
Designed for current and
foreseeable requirements
Refactoring Inexpensive Expensive
Size Smaller teams and products Larger teams and products
Primary objective Rapid value High assurance
TABLE II
MAPPING AGILE HOME GROUND THEMES TO TRAITS OBSERVED IN MOBILE
SOFTWARE DEVELOPMENT (ADAPTED FROM [20])

Ideal Agile
Characteristic
Rationale Mobile software
High
environment
volatility
Due to high change of
requirements, less need for up-
front design & planning, need
for an incremental and iterative
development approach.
High uncertainty, dynamic
environment: Hundreds of new
mobile phones produced each year
Small
development
teams
Small teams are able to react
more rapidly, share
information, need less
documentation , etc.
Majority of mobile software is
developed in micro or SME
companies, or development teams.
Identifiable
customer
To avoid business
misunderstanding
Potentially unlimited number of
end-users. Business customer
easier to identify, e.g. distributor.
Object-oriented
development
environment
Most tools that support agile
development exist for object
oriented platforms.
E.g., Java and C++ used; some
problems in proper tooling e.g. for
refactoring and test-first approach
Non-safety
critical software
Failures do not cause loss of
lives. More agility can be
pursued.
Majority of existing mobile
software is for entertainment
purposes. Mobile terminals are not
reliable.
Application-
level software
Large embedded systems
require extensive
communication & verification
mechanisms.
While mobile systems are
complex and highly dependent,
mobile applications can be stand-
alone applications.
Small systems Less upfront design needed
Mobile applications vary in size,
but are generally less than 10000
lines of code.
Short
development
cycles
For the purposes of rapid
feedback
Development cycles vary. Typical
mobile applications and services
can be developed within a 1-6
month time-frame.
D. Architecture-Based Development
The efficiency of the software product line approach
depends on the development firms ability to invest in the
development of a common platform. This necessitates the
development of a general architecture for such products.
E. Support for Reusability
The need for using functionally equivalent components
(such as message boxes in a specific language), and
functionally similar components (such as security operation
classes), along with tight time constraints on the development
of mobile software, requires the developer to use reusable
components extensively. Having to develop these components
from scratch each time they are needed increases the cost of
mobile software development, delays product delivery, and
makes the software more error prone. Support for component-
based and layer-based development approaches is therefore
essential in a mobile software development methodology.
F. Inclusion of Review and Learning Sessions
One of the most effective factors in a production companys
success is its ability to abstract the knowledge obtained during
product development. As mentioned above, todays mobile
software development industry tends to be product-oriented;
the methodology should therefore incorporate review sessions
throughout the process to ensure product analysis, and include
lessons-learnt sessions after delivering a working product to
the market to ensure that experiences are analyzed and logged.
G. Early Specification of Physical Architecture
Mobile terminals constraints should be considered from
very early stages of software design. In fact, a high degree of
technical risk can be traced to mobile hosts limitations and
their differences in the implementation of basic features.
Consequently, the physical architecture should be elaborated
in the early stages of software development. A prototype may
also be required in order to mitigate technical risk.
V. BUILDING A MOBILE SOFTWARE DEVELOPMENT
METHODOLOGY USING METHOD ENGINEERING TECHNIQUES
Motivated by the belief that no one methodology fits all
situations, Methodology Engineering was first introduced as a
discipline aimed at constructing methodologies to match given
organizational settings or specific development projects [22].
The discipline later came to be known as Method Engineering,
with the definition broadened as: The engineering discipline
to design, construct, and adapt methods, techniques and tools
for the development of information systems [3].
There are several approaches to method engineering [23]:
Ad-hoc: Constructing a new methodology from scratch;
Paradigm-based: Instantiating, abstracting or adapting an
existing meta-model to produce the target methodology;
Extension-based: Enhancing an existing methodology
with new concepts and properties;
Assembly-based: Constructing the methodology through
assembling method fragments retrieved from a repository.
In this section, we describe our work in utilizing an
approach called Hybrid Methodology Design [4], through
which the target methodology has been built based on the
requirements defined in the previous section and knowledge
gathered from existing methodologies and process
patterns/metamodels [4], [6], [24]. The Hybrid Methodology
Design process has been devised as a top-down iterative-
incremental process consisting of the following tasks [4]:
Prioritization of the Requirements: performed at the start
of the process and repeated at the end of each iteration.
The requirements are ordered according to their relevance
to the current scope and level of abstraction, focusing the
design process on satisfying requirements of higher
significance. At the start of the process, abstraction is at
its highest level and the scope encompasses the whole
lifecycle, therefore requirements with lifecycle-level
impacts are given precedence; as design progresses to
lower levels of abstraction, priority is gradually shifted to
requirements with finer-grained aspects.
Iterative Design Engine: The following tasks are
performed in each iteration:
o Selection of the design approaches to be used in
the current iteration: The possible approaches to
designing the target methodology include:
Instantiation: instantiating an already
available process metamodel;
Artifact-oriented: devising a seamless
complementary chain of artifacts and
building the process around it;
Composition: using one of the already
available libraries of process patterns;
Integration: integrating features, ideas and
techniques from existing methodologies.
o Application of the selected design approaches
aimed at defining the methodology at the current
scope and level of abstraction: Special attention
should be given to existing methodologies and
process patterns/metamodels, thus implementing
features of strength and avoiding common pitfalls.
The prioritized set of requirements focuses the
design effort on satisfying requirements of
importance. The methodology elements designed
are then integrated into the produced blueprint.
o Revision, refinement and restructuring of the
methodology built so far in order to accommodate
the changes made in the current iteration.
o Specification of the level of abstraction for the
next iteration, and definition of the scope and
intended level of detail.
o Revision and refinement of the requirements,
including prioritization according to the scope and
level of abstraction intended for the next iteration.
The iterative-incremental engine at the core of the design
process generates the methodology in a top-down fashion
from the general lifecycle to the details of activities using
the requirements and methodology descriptions as a basis.
Of the four design approaches used in each iteration, two
i.e. Integration and Artifact-oriented are relatively novel in
this context. The Integration approach promotes integrating
ideas and techniques directly from existing methodologies,
instead of first dissecting the methodologies into fragments (as
is common practice in assembly-based method engineering);
the intention is to preserve synergy and avoid loss of
functional capacity. Different approaches of the Hybrid
approach have different uses depending on the scope and
abstraction level of the design activity undertaken in the
current iteration: Instantiation is more useful when designing
high-level aspects of the methodology, Integration and
Composition are more suited to the design needs of low-level
aspects, and the Artifact-oriented approach comes in between,
i.e. while less useful at the general lifecycle level, it is
indispensable when addressing seamlessness at the inter-
subprocess and intra-subprocess levels. We constructed the
overall framework of the target methodology through four
iterations of the Hybrid Design Engine, with the results shown
in Figure 1. In following the Instantiation approach, the
generic software development life cycle was set as the base
process, i.e. at the highest level of abstraction. It provides
adequate high-level coverage of generic software development
activities and supports umbrella activities, thus providing the
framework needed for satisfying the requirements through
further iterations of the engine.
In the first iteration, the methodology was elaborated via the
use of generic patterns for risk-based, architecture-centric, and
test-based development. Analysis was split into Preliminary
Analysis and Detailed Analysis, in order to mitigate
development risks. To achieve architecture-based
development, Design was split into Architectural Design and
Detailed Design, with the relevant feasibility analysis,
planning and architectural design activities duly added. The
Implementation and Test sub-processes were combined in
order to accommodate test-based development. These patterns
are frequently used in agile methodologies. Moreover, since
the main focus in current mobile software development is on
product development, we incorporate the Commercialization
phase; this would also facilitate product line support. The
design approach in this iteration was mainly instantiation,
using meta-models including SPEM [25] and OPF [26]
and general object-oriented lifecycles such as OOSP [27].
In the second iteration, market consciousness was focused
upon through borrowing activities from New Product
Fig. 1. Gradual refinement of our proposed methodology during iterations of the Hybrid Methodology Design process
Development. These include Idea Generation (incorporated
into the beginning phases of the process), and Market Testing
(performed before commercializing the software). The design
approach in the second iteration was integration, using
reusable parts of the NPD process.
In the third iteration, we enhanced our processs
development engine by incorporating ideas from Adaptive
Software Development (ASD) [28]. ASD is a component-
based agile methodology especially rich in quality assurance
measures. The Speculate-Collaborate-Learn cycle that forms
the basis of the ASD process provides the means to cope with
the uncertainties of software development. Using the
integration approach, the iteration resulted in the
incorporation of review sessions into the methodology. By
benefiting from component-based development practices, we
have also improved our processs support for reusability.
Considering the potential technology risks in mobile
software development, the last iteration was mainly concerned
with adding prototyping (as a process pattern [27]) to our
process. Prototyping also facilitates the early specification of
the physical architecture. Moreover, the process was refined
through moving the Preliminary and Business Analysis
activity into Project Initiation, with output artifacts based on
those produced by ASDs Project Initiation phase. The design
approach used was mainly composition, with integration used
when reusing ideas from the ASD methodology.
VI. CONCLUSIONS
We have identified the main requirements of a mobile
software development methodology, based on which a high-
level methodology framework was built using the Hybrid
Methodology Design approach. Our proposed methodology is
an agile risk-based methodology, highly influenced by the
ASD method and NPD approaches. The requirements-based
nature of the Hybrid approach ensures that the requirements
are properly addressed, and validating the resulting
methodology against the requirements seems to confirm this.
In furthering this research, our next step would be to apply
further iterations of the Hybrid Design Engine at lower levels
of abstraction, thereby specifying the finer-grained tasks of the
process. The methodology can then be put to test in
developing commercial mobile software products; the process,
and the requirements on which it is based, can thus be tuned.
Another strand of research can focus on extending our work to
the development of mobile software projects, rather than
products; since technology advances with incredible speed,
this is likely to be an industry demand in the future.
REFERENCES
[1] G. H. Forman, J. Zahorjan, The Challenges of Mobile Computing.
IEEE Computer, Vol. 27, No. 4, April 1994, pp. 38-47.
[2] I. S. Heyes, Just Enough Wireless Computing, Prentice Hall, 2002.
[3] S. Brinkkemper, Method engineering: engineering of information
systems development methods and tools, Information and Software
Technology, Vol. 38, No. 4, 1996, pp. 275-280.
[4] R. Ramsin, The Engineering of an Object-Oriented Software
Engineering Methodology. Ph.D. Thesis, University of York, York,
UK, 2006. Available: http://www.cs.york.ac.uk/ftpdir/reports/YCST-
2006-12.pdf.
[5] J. Highsmith, Adaptive Software Development: A Collaborative
Approach to Managing Complex Systems. Dorset House, 2000.
[6] K.T. Ulrich, S.D. Eppinger, Product Design and Development, 3rd
Edition, McGraw-Hill, 2004.
[7] M. Satyanarayanan, Fundamental Challenges in Mobile Computing, in
Proc. of the Fifteenth annual ACM symposium on Principles of
distributed computing, 1996, pp. 1-7.
[8] M. Satyanarayanan, Pervasive Computing: Vision and Challenges,
IEEE Personal Communications, August 2001.
[9] A. Agostini, C. Bettini, N. Cesa-Bianchi, D. Maggiorini, D. Riboni, M.
Ruberi, C. Sala, and D. Vitali, Towards Highly Adaptive Services for
Mobile Systems, in Proc. of the Working Conference on Mobile
Information Systems (IFIP TC8), 2004.
[10] H. Grine, T. Delot, S. Lecomte, Adaptive Query Processing in Mobile
Environments, in Proc. of the 3rd international workshop on
Middleware, 2005.
[11] V. Pham, A. Karmouch, Mobile Software Agents: An Overview,
IEEE Communications Magazine, July 1998.
[12] E. Valavanis, C. Ververidis, M. Vazirgianis, G.C. Polyzos, K. Norvag,
MobiShare, Sharing Context-Dependent Data & Services from Mobile
Sources, in Proc. of the IEEE/WIC International Conference on Web
Intelligence (WI 2003), 2003, pp. 263-270.
[13] M. Haahr, R. Cunningham, V. Cahill, Towards a Generic Architecture
for Mobile Object-Oriented Applications, in Proc. of the 2000 IEEE
Workshop on Service Portability and Virtual Customer Environments,
2000, pp. 91-96.
[14] E.S. Chung, J.I. Hong, J. Lin, M.K. Prabaker, J.A. Landay, and A.L. Liu,
Development and Evaluation of Emerging Design Patterns for
Ubiquitous Computing, in Proc. of the 2004 conference on Designing
interactive systems: processes, practices, methods, and techniques,
2004, pp. 233-242.
[15] O. Gerstheimer and C. Lupp, Needs versus technology - the challenge
to design third generation mobile applications, Journal of Business
Research, Vol. 57, No. 12, Dec. 2004.
[16] A.M. Vainio, T. Tuunanen, P. Abrahamsson, Developing Software
Products for Mobile Markets: Need for Rethinking Development Models
and Practice, in Proc. of the 38th Hawaii International Conference on
System Sciences (HICSS'05), Jan. 2005.
[17] P. Abrahamsson, A. Hanhineva, H. Hulkko, T. Ihme, J. Jlinoja, M.
Korkala, J. Koskela, P. Kyllnen, and O. Salo, Mobile-D: An Agile
Approach for Mobile Application Development, in Proc of the
OOPSLA04 Conference, 2004.
[18] B. Boehm and R. Turner, Balancing agility and discipline: A guide for
the perplexed, Addison-Wesley, 2003.
[19] B. Boehm, Get ready for agile methods, with care, IEEE Computer,
Vol. 35, No. 1, 2002, pp. 64-69.
[20] P. Abrahamsson, Keynote: Mobile software development the
business opportunity of today, in Proc. of the International Conference
on Software Development, 2005, pp. 20-23.
[21] P. Clements, L. Northrop, "Software Product Lines", course notes of
Product Line Systems Program, Software Engineering Institute,
Carnegie Mellon University, 2003.
[22] K. Kumar, R. J. Welke, Method Engineering: a proposal for situation-
specific methodology construction, in Systems Analysis and Design: A
Research Agenda, 1992.
[23] J. Ralyt, R. Deneckre, C. Rolland, Towards a generic model for
situational method engineering, in Proc. of CAiSE03 (LNCS 2681),
2003, pp. 95-110.
[24] R. Ramsin, R. F. Paige, "Process-centered review of object-oriented
software development methodologies", ACM Computing Surveys, Vol.
40, No. 1 (February), 2008, Article 3, pp.1-89.
[25] OMG, Software Process Engineering Metamodel Specification (v1.0),
Object Management Group (OMG), 2002.
[26] D. Firesmith, B. Henderson-sellers, The OPEN Process Framework: An
Introduction, Addison-Wesley, 2001.
[27] S. W. Ambler, Process Patterns: Building Large-Scale Systems Using
Object Technology, Cambridge University Press, 1998.
[28] J. Highsmith, Messy, exciting, and anxiety-ridden: Adaptive software
development, American Programmer, Vol. 10, No. 4 (April), 1997, pp.
23-29.