An introduction to software

development
About this free course
This free course is an adapted extract from the Open University course M813 Software development
http://www.open.ac.uk/postgraduate/modules/m813.
This version of the content may include video, images and interactive content that may not be optimised
for your device.
You can experience this free course as it was originally designed on OpenLearn, the home of free
learning from The Open University:
www.open.edu/openlearn/science-maths-technology/introduction-software-development/content-sec-
tion-0.
There you’ll also be able to track your progress via your activity record, which you can use to
demonstrate your learning.
Copyright © 2016 The Open University
Intellectual property
Unless otherwise stated, this resource is released under the terms of the Creative Commons Licence
v4.0 http://creativecommons.org/licenses/by-nc-sa/4.0/deed.en_GB. Within that The Open University
interprets this licence in the following way:
www.open.edu/openlearn/about-openlearn/frequently-asked-questions-on-openlearn. Copyright and
rights falling outside the terms of the Creative Commons Licence are retained or controlled by The Open
University. Please read the full text before using any of the content.
We believe the primary barrier to accessing high-quality educational experiences is cost, which is why
we aim to publish as much free content as possible under an open licence. If it proves difficult to release
content under our preferred Creative Commons licence (e.g. because we can’t afford or gain the
clearances or find suitable alternatives), we will still release the materials for free under a personal end-
user licence.
This is because the learning experience will always be the same high quality offering and that should
always be seen as positive – even if at times the licensing is different to Creative Commons.
When using the content you must attribute us (The Open University) (the OU) and any identified author in
accordance with the terms of the Creative Commons Licence.
The Acknowledgements section is used to list, amongst other things, third party (Proprietary), licensed
content which is not subject to Creative Commons licensing. Proprietary content must be used (retained)
intact and in context to the content at all times.
The Acknowledgements section is also used to bring to your attention any other Special Restrictions
which may apply to the content. For example there may be times when the Creative Commons Non-
Commercial Sharealike licence does not apply to any of the content even if owned by us (The Open
University). In these instances, unless stated otherwise, the content may be used for personal and non-
commercial use.
We have also identified as Proprietary other material included in the content which is not subject to
Creative Commons Licence. These are OU logos, trading names and may extend to certain
photographic and video images and sound recordings and any other material as may be brought to your
attention.
Unauthorised use of any of the content may constitute a breach of the terms and conditions and/or
intellectual property laws.
We reserve the right to alter, amend or bring to an end any terms and conditions provided here without
notice.
All rights falling outside the terms of the Creative Commons licence are retained or controlled by The
Open University.
Head of Intellectual Property, The Open University

2 of 39 Wednesday 6 November 2019

Contents
Introduction 4
Learning Outcomes 5
1 Software development as engineering 6
2 Software development processes 7
3 Why is software development difficult? 10
3.1 A philosophical perspective 10
3.2 A historical perspective 13
4 Risk 15
5 Software quality 16
6 Modelling and the UML 19
7 Object orientation 22
7.1 Modelling with objects 22
7.2 Programming with objects 23
8 Finding and reading academic articles 25
8.1 A workflow for reading the academic literature 25
Conclusion 37
References 37
Acknowledgements 38

3 of 39 Wednesday 6 November 2019

Introduction

Introduction
This OpenLearn free course is about software development, the discipline concerned with
the methods, techniques and processes of building software artefacts: today's software
development professional will need an in-depth knowledge of this discipline to be able to
create software which best serves the needs of our digital economy and society. Software
development is also a fast-moving discipline, so that the software development
professional also needs effective strategies for finding materials that track its leading
edge.
This OpenLearn course is an adapted extract from the Open University course
M813 Software development.

4 of 39 Wednesday 6 November 2019

Learning Outcomes
After studying this course, you should be able to:
● appreciate the engineering nature of software development
● describe key activities in software development and the role of modelling
● explain key concepts in software development such as risk and quality
● explain the basics of an object-oriented approach to software development
● describe a simple workflow for interacting with the published literature on software development.
1 Software development as engineering

1 Software development as engineering

G. F. C. Rogers, writing in the early 1980s, defined engineering as ‘the practice of
organising the design and construction of any artifice which transforms the physical world
around us to meet some recognised need’ (Rogers, 1983).
Rogers was not a software developer, but an engineer specialising in thermodynamics
and jet-engine development. Yet his definition captures the essence of engineering and
has a wide applicability which extends to software. Much of our endeavour in software
development is the design and construction of software to meet some recognised need –
of people, organisations or society at large – with tangible effect on the real world.

Activity 1
Look at Rogers’ definition of engineering. How would you relate elements of that
definition to your understanding of software development?
Discussion
Organising design and construction can be related to what is commonly known as a
‘software development process’: if you develop software in the context of a commercial
organisation, it is highly likely that you and your team will follow one such process to
coordinate the construction and deployment of software.
The physical world around us can be related to the context in which the software will
be deployed – the context could be the way an information system is organised, the
device in the case of a software controller, or the internet for a web application.
The recognised need in Rogers’ definition relates to the problems the software is
meant to solve – what we commonly capture as requirements in software
development. These problems are the reason why the software should exist, and
express how software should affect its context once in operation.
In this course we take an engineering view of software development, and, guided by
Rogers’ definition, will explore in some detail the constituent elements of software
development, and the interrelations between these elements.

6 of 39 Wednesday 6 November 2019

2 Software development processes

2 Software development processes

With reference to Rogers’ definition, a software development process is the practice of
organising the design and construction of software and its deployment in context.
Effective software development processes remain the holy grail of software development
and over the years many contenders have emerged and then gone out of fashion. These
vary in the detail of what activities they prescribe, their related artefacts, how the activities
should be carried out in relation to one another, and how often each activity should be
revisited. There are, however, some fundamental development activities that are well
understood and common to most approaches. In very broad terms:

● Analysis involves understanding the problem which the software is intended to

solve, i.e., the requirements in context, with validation as the means to check that
understanding.
● Design involves describing, conceptually, a software solution that meets the
requirements of the problem.
● Implementation involves realising such a solution in software.
● Testing involves making sure that the solution has certain inherent qualities, with
verification as a means to check its adequacy with respect to the specified
requirements and validation as a means to check that the solution does address the
problem.
● Deployment involves making the developed solution available in its context of
operation and use.

Note: validation and verification are sometimes confused. One way to understand
the distinction is to think of validation as addressing the question ‘are we solving the
right problem?’ and verification as addressing the question ‘are we solving the problem
right?’

Activity 2
If requirements are validated against the problem, and the implemented system is
verified against those requirements, then the system should indeed be addressing the
problem. Why then do we still need to validate the system against the problem during
testing?
Discussion
The core of the problem is that the system needs to satisfy the requirements in its
context of operation. During software design and construction we rely on assumptions
about the properties of that context – which might turn out to be inaccurate once the
system is in operation, or we may discover that there are other critical properties which
we have overlooked. The context of operation of a software system is often a complex
place whose characteristics and behaviour are difficult to grasp and predict completely
and accurately, even with the best engineering effort.

7 of 39 Wednesday 6 November 2019

2 Software development processes

While this course will focus primarily on software development activities, it is important to
remember that the process does not end with a deployed product. The full software life-
cycle also includes:

● maintenance: makes sure that the deployed solution continues to address the
problem throughout the time it is in operation, and handles its decommissioning at
the end of the life-cycle.

Note: although software maintenance is a well understood term, some authors prefer to
talk about software evolution, meaning that software continually changes over its life-
cycle, with changes often triggered by changes in context and in stakeholders’ needs.
As a consequence, software development activities typically carry on post-deployment
as new releases of the software are developed and deployed.

Activity 3
Do a web search for ‘software development processes’. Name three such processes
that recur in your search results. What are their key characteristics, and differences
with reference to the activities mentioned above?

In completing Activity 3 you may have encountered some of the following.

The waterfall development process model (or simply, waterfall model; Royce,1970;
Benington, 1983) was inspired by manufacturing processes. Historically, this was the
earliest software development process to be widely recognised and applied. It relies on
the definition of sequential phases of analysis, design, implementation and testing, each
starting only after the previous one has finished; in its purest form, all the analysis is done
first, followed by all the design, then the implementation, and finally the testing. The
waterfall process has been widely used and has yielded many successful software
products, and variants of it (e.g. the V-model; see Forsberg and Mooz, 1992) are still in
use in some industries, particularly in highly regulated sectors. However, the waterfall
process is based on several assumptions that have been questioned. For example, it
relies on the existence of a set of requirements that is defined beforehand and remains
unchanged during development; it also assumes that all code is designed from scratch,
making no allowance for the reuse of existing software. Because the development is
carried out as a linear process, the waterfall process does not allow previous phases of
the development to be repeated without repeating the whole sequence of steps.
Iterative and incremental processes emerged as a reaction to the recognised
limitations of the waterfall process, particularly the latter’s rigidity in the face of change
and its lack of support for reuse. With iterative and incremental processes there is still a
need for analysis, design, implementation and testing activities, but these activities are
carried out in a more flexible way than in the waterfall process. Many iterative and
incremental processes exist – for example, the spiral model (Boehm, 1988) or the Unified
Process (Jacobson et al., 1999) – and all share some basic features: they are ‘iterative’ in
the sense that one or more activities are repeated; and they are ‘incremental’ in the sense
that software development proceeds from an initial subset of the requirements to more
and more complete subsets, until the whole system is addressed. Hence, an iterative and
incremental process consists of several cycles of analysis, design, implementation and
testing, each providing feedback for the next cycle, in which evermore refined and
enhanced levels of development are achieved.

8 of 39 Wednesday 6 November 2019

2 Software development processes

A popular family of iterative and incremental processes emerged at the start of this
century, covered by the umbrella term of Agile development (Cockburn and Highsmith,
2001; Fowler and Highsmith, 2001). These processes encourage continual realignment of
development goals with the needs and expectations of the customer, and provide ‘lighter-
weight’, faster and ‘nimbler’ software development processes that can adapt to the
inevitable changes in customer requirements. Hence, such processes promote frequent,
small, incremental cycles of analysis, design, implementation and testing. Also, working
code is valued above other types of artefact, such as models or other documentation.
Extreme Programming (XP) (Beck, 2000) and Scrum (Schwaber and Beedle, 2001) are
two well-known Agile development approaches.
Within the same timescale of Agile development, model-driven-development (Se-
lic, 2003) also arose. This approach sees models and not code as the key artefacts in
software development. According to this approach, models allow the developer to work at
a much higher level of abstraction than code, and hence be closer to the business and its
processes; also, models are platform-independent and hence decoupled from specific
technology. Well defined syntax (how models should be defined) and semantics (how they
should be interpreted) allow code to be derived from models through transformations and
a suite of automated tools are being developed to support the approach. With model-
driven development, the boundary between analysis and design becomes blurred, with
development progressing seamlessly through transformed models.

9 of 39 Wednesday 6 November 2019

3 Why is software development difficult?

3 Why is software development difficult?

You may be wondering why so many development processes have been proposed, why
they are the way they are, or even why we need software development processes at all.
The simple answer is that developing software is hard and development processes are
regarded as a way to manage the risk of getting it wrong: in 2004 the total cost of
information systems failure across the European Union was an estimated €142 billion,
while McManus and Wood-Harper’s 2008 study of 214 large IT projects found that 23.8%
were cancelled before delivery, and of those that ran to completion, 42% overran in terms
of time and/or cost. Similar figures are cited in the CHAOS Summary (Standish
Group, 2009), which reported that ‘32% of all projects succeeding (delivered on time, on
budget, with required features and functions); 44% were challenged (late, over budget,
and/or with less than the required features and functions); and 24% failed (cancelled prior
to completion or delivered and never used)’.
So, why is software development so difficult? This is a topic that has exercised software
thinkers and practitioners alike for many decades.

3.1 A philosophical perspective

In the late 1980s W. M. Turski (1986) observed that developing software is inherently
difficult because it involves descriptions of properties of both formal and non-formal
domains. Formal domains are those which pertain to the artificial, i.e., the software itself
and its hardware – man-made artefacts which can be described using fully formal
notations, such as code or statements of logic. Non-formal domains are those which
pertain to the natural, such as people, organisations or natural phenomena; such domains
don’t always allow formal description, and even when they do, different linguistic
descriptions are used, which raises the further challenge of how to relate them and reason
about them. Consequently, Turski sees developing software as compounding the
difficulties of mathematics, in its concern with relationships between formal domains, and
of natural science, in its concern with descriptions of the properties of non-formal
domains. In Turski’s view an effective software development process would have to deal
adequately with descriptions across the formal/non-formal divide.

Activity 4
Think of various descriptions used in software development. Which do you think
pertain to the formal and non-formal domains identified by Turski?
Discussion
You may have considered some of the following.
Code is intrinsically a formal description, usually expressed in a language with formally
specified syntax and semantics. Software design can also be captured formally,
although semi-formal descriptions (e.g., sketches) can also be used to describe it.
Requirements statements are usually non-formal descriptions, often expressed in
natural language, that try to capture properties of formal and informal domains,
e.g., how people might interact with the system and how the system might respond in
turn. Some requirements can be specified using formal languages and notations; this
is particularly true in certain domains, as with embedded safety-critical software,

10 of 39 Wednesday 6 November 2019

3 Why is software development difficult?

although the limitations of formal language for expressing general requirements are
widely acknowledged.

Building upon Turski’s thinking, M. Jackson and others (Jackson, 1995; Gunter et al.,
2000; Hall and Rapanotti, 2003) proposed a view of software development encompassing
two distinct but related ‘spaces’: the problem space, in which the problem to be
addressed is located; and the solution space, in which the software solution resides. In
this view, descriptions, whether formal or non-formal, are further categorised into
indicative and optative: an indicative description is a statement of fact – that is, an
assumption which is believed to be true and on which further reasoning can be
constructed; an optative description is a statement of wish – that is, something we would
like to become true at some point. To give an example, a requirement is an optative
description, as it states something that it is hoped will be true once the system is designed
and operational; on the other hand, an existing business rule is an indicative description,
as it states how the business works currently; that is, this is something taken for granted in
developing the system.

Note: one way to understand the distinction between optative and indicative is to
consider optative descriptions as statements of something we wish to be true in the
future, but which may not be true now, and indicative descriptions as statements of
something which is true now and which we expect still to be true in the future.

In this view, specifications provide the point of contact between the two spaces: a
specification captures optative requirements as to how the software system should
behave at its interface with the real world. As such, the specification can be validated
against the problem: we can check whether building a software system to that
specification will address the problem; and the software can be verified against the
specification: i.e., we can check that the software has been built in such a way that it will
satisfy the specification. In this view, therefore, an effective software development
process would have to deal adequately with building and separating optative and
indicative descriptions across the problem/solution divide, and transform optative
requirements before development into indicative statements once the software is in
operation.

Activity 5
Consider the problem and solution spaces proposed by Jackson and others. How
would you describe the way the waterfall process spans the two? What about iterative
and incremental development processes?
Discussion
The waterfall model assumes that we proceed in linear fashion from the problem to the
solution space: first, through analysis, we understand the problem and express
requirements and specification; then we proceed to design and implementation in the
solution space, and test the outcome against the specification.
On the other hand, iterative and incremental processes assume that we cross the
problem/solution divide many times: at each iteration we revisit the problem space to
identify new requirements and define new specifications; we then proceed to the
solution space for related design, implementation and testing.

11 of 39 Wednesday 6 November 2019

3 Why is software development difficult?

The work of Turski and of Jackson and others provides useful insights into the nature of
the descriptions required in software development, and processes can be, and have
been, defined which establish when and where such descriptions are required. However,
there are further dimensions to software development which are widely recognised as
being highly influential, and which should be taken into account. One such dimension in
those descriptions is fitness for purpose – and how this can be established. The other
dimension is the design rationale, i.e., the explicit articulation and capture of the reasons
behind and justifications for decisions made when designing an artefact, including
alternatives considered and trade-offs.
Hall and Rapanotti (2009) have proposed a view of software development as spanning
two orthogonal spaces: the design space, to which the descriptions of Turski and
Jackson and others belong; and the assurance space, where consideration of design
rationale and fitness for purpose take place and related descriptions are generated. In this
view, contact between the two spaces occurs throughout the development process, both
in the problem and solution spaces, at points at which assurance artefacts are shared and
checked by different stakeholders – for example, as with a business analyst checking
understanding of business processes with the customer, or a software architect reviewing
and approving an architectural blueprint for a system. Hence, the fitness for purpose of
any description, from problem statement to working code, can only be established by the
stakeholders involved in the process, bringing stakeholders to the heart of software
development and highlighting their identification and participation in the process as a key
success factor. As a consequence, in this view, an effective software development
process would have to deal adequately with building and separating descriptions not just
across the problem/solution divide, but also across the design/assurance divide, and
would also have to identify key representative stakeholders and account for their
involvement in the development process.

Activity 6
Consider Hall and Rapanotti’s design and assurance spaces. To what extent does the
waterfall process address the relationship between these two spaces? What about
iterative and incremental development processes?
Discussion
The waterfall model fits primarily within the design space, as it makes no explicit
reference to assurance and its relevant stakeholders. The only activity that could be
related to the assurance space is testing, which can be interpreted as developing
assurance descriptions – the set of test cases – to demonstrate specific properties of
the working software, often to the customer. Similar observations can be made of
iterative and incremental processes, which have developed directly from the waterfall
model.

Historically, software development processes have focused primarily on the design space.
However, the importance of the assurance space has long been acknowledged and
recognised, and a parallel discipline of project management has developed in a bid to
address some of the assurance needs of software development processes highlighted by
Hall and Rapanotti (2009). It should be noted that the importance of stakeholders is also
explicitly recognised by Agile development, whose manifesto explicitly values individuals
and interactions as well as customer collaboration.

12 of 39 Wednesday 6 November 2019

3 Why is software development difficult?

3.2 A historical perspective

Manchester University was the home of the world’s first stored-program computer, the
Small-Scale Experimental Machine – also known as ‘Baby’ – which ran its first program on
21 June 1948 (Tootill, 1998). Baby’s first run was on a 17-line program that calculated the
highest factor of an integer. This was not a mathematical problem of much significance,
but was chosen as a proof-of-concept because it made use of all seven instructions in
Baby’s instruction set (Enticknap, 1998).
This trivial problem was, arguably, the beginning of the modern information age. At that
time, a simple software process was sufficient: the program was simply crafted by
inspection of the problem and debugged until it was correct. Things didn’t stay that simple
for long. By the early 1990s, problem complexity had increased significantly, with code
complexity increasing in turn by several orders of magnitude; for instance, by 1991 the
IBM CICS (Customer Information Control System) had grown to 800,000 lines of code.
A first response to the need to manage higher code complexity was to provide richer and
richer structures within the programming languages in which to implement that code.
High-level programming languages were developed to provide greater abstraction, often
with concepts from the problem domain allowing more complex code to service more
complex problems. There was a steady trend in sophistication, leading to the
programming languages we still make use of today. Then the concept of software
architecture was developed, placing at the heart of design repeated code and thought
structures, which were independent of programming language and could be used to
structure large systems (Bass et al., 2003).
Coincident with the rise of code complexity was a growing realisation of the importance of
an engineering approach to software; here, the traditional engineering disciplines inspired
early software development processes, as with the waterfall model. Also, with architecture
at the heart of code design, problem complexity started to be addressed through software
requirements engineering – see, for instance, Robertson and Robertson (2006). This
complemented code design through phases of elicitation, modelling, analysis, validation
(i.e., ‘are we addressing the correct problem?’) and verification (i.e., ‘have we addressed
the problem correctly?’). Now, software requirements engineering led from system
requirements to software requirements specifications on which architectural analysis
could be performed.
Things did not go smoothly with early development processes. Already in 1970 Royce
saw the need – based on his extensive experience in developing embedded software – to
iterate between software requirements and program design (and between program
design and testing). For software, the early stages of development could not always be
completed before commencing the later stages. Indeed, it was Royce’s enhanced
waterfall model, able to cope with requirements incompleteness, that inspired a
generation of iterative and incremental software processes (Royce, 1970).
The waterfall model also assumed that a stable problem exists on the basis of which to
begin development. This proved to be adequate in early applications. For instance, a very
early business use of computers came about in 1951, when the Lyons Electronic Office
(LEO) was first used for order control and payroll administration (Bird, 1994). Early
applications typically involved scheduling, reporting and ‘number-crunching’, and
although they provided many technical challenges that were complex for that time, what
amounts to a first-mover advantage (Grant, 2007) meant the Lyons company enjoyed a
stable environment in which to explore its use of computers. However, the drive for more
sophisticated business computing meant that developers now had to deal with the

13 of 39 Wednesday 6 November 2019

3 Why is software development difficult?

volatility of the business world and the way this affected its relationship with the formal
world of the computer. Therefore, the waterfall assumption soon showed itself to be
inadequate for many real-world applications.
Volatility also arises from the nature of technology: rapid technological change also drives
changes in the context in which software is deployed, and any connection between the
informal world outside the computer and the formal world of the computer must link two
volatile targets. Due to volatility, the resulting relationship between requirements and
architectures is an order of magnitude more complex than for stable systems, and the
treatment of it offered by traditional techniques, such as those derived from Royce’s
processes, are increasingly seen as inadequate. Indeed, such volatility is now recognised
as one of the greatest challenges software development faces.

14 of 39 Wednesday 6 November 2019

4 Risk

4 Risk
As we have discussed, complexity and volatility are a cause of difficulty in software
development and increase the risk of failure: failure to deliver in time and on budget, or to
meet stakeholders’ needs in context. In turn, failure can have profound consequences: it
may lead to tangible harm or losses (e.g. harm to people or damage to goods, loss of
contracts, loss of revenue, decrease in market share, penalties in contracts) or intangible
harm (e.g. loss of trust, credibility or future business opportunities, damage to the
reputations of people, organisations or trademarks, dissemination of confidential
information, loss of intellectual-property rights).

Note: risk assessment (how to identify risk) and risk management (how to deal with
risk) have been the subject of study in a wide variety of domains, from business to
engineering, health or statistics, and a vast literature exists on the subject.

One important role of software development processes is to reduce risk, particularly in

relation to two fundamental categories:

● Problem-related risks: these are risks associated with misunderstanding the problem
and the stakeholders’ requirements, or making inappropriate assumptions as to the
context of the system.
● Solution-related risks: these are factors associated with developing software which
does not address the identified problem or is not fit for its purpose.

Activity 7
Do a web search on software development risk. Which recurrent factors can you
identify? How do they relate to the categories above?
Discussion
Your list will be different from ours. Among the recurrent risk factors we identified are:

● Problem-related risks: incomplete or inconsistent requirements; unachievable

goals; wrong assumptions about the domain.
● Solution-related risks: technical infeasibility; breakdown of specification.

15 of 39 Wednesday 6 November 2019

5 Software quality

5 Software quality
In very abstract terms, the quality of an artefact, whether it is software or otherwise,
relates to the presence of positive characteristics of the artefact which will satisfy
stakeholders’ needs and expectations in a particular real-world context. Quality also
relates to the absence of negative characteristics which may harm or frustrate
stakeholders and fail to meet their needs and expectations in that context. It follows that
quality is not an absolute property: it can only be judged and measured with reference to
stakeholders’ needs and expectations within their context. This is particularly true of
software, which is more often than not developed to address real-world problems as part
of much wider systems, usually involving people, processes and other forms of
technology.
So quality has to do with fitness for purpose, and the hallmark of a good quality product
is that its diverse stakeholders are satisfied with it both now and in the foreseeable future.
Managing software quality is then a continuous process aimed at ensuring that the
software’s fitness for purpose is maintained throughout the software’s life-cycle.
Failure to attain fitness for purpose can have disastrous consequences, as witnessed by
the many examples of software failure which have been reported over the years.

Activity 8
Do a web search on software failures. What did you find?
Discussion
You may have found plenty of high-profile software failures reported by the media and
in the academic literature. We found a long list of postings and articles. For instance,
Lake (2010) summarised 11 ‘epic’ software failures attributable to deliberate decisions
by programmers (rather than, say, faults or malicious attacks), in fields ranging from
space exploration and telecommunications systems, in which the loss was mainly
financial, to medical and aviation applications, in which the result was a catastrophic
loss of lives.

Activity 9
What do you think the quality expectations for an air traffic control software system
might be? What about those for a personal diary mobile app?
Discussion
As lives depend on it, an air traffic control system will need to be highly dependable
and safe to operate, and provide accurate and reliable information in real time which
can be readily interpreted and understood by highly trained people – particularly the air
traffic controllers who need to direct pilots. Lack of integrity of data or unavailability of
the system could have catastrophic consequences.
Some aspects of dependability – such as data integrity and reliability – would also be
desirable but not nearly as critical in a personal diary app. Ease of use would also be
desirable as the app is likely to be used by people with all levels of skills.

16 of 39 Wednesday 6 November 2019

5 Software quality

While quality expectations might be different in different domains, there is a common need
to build quality into product design, to increase the likelihood that the product is fit for
purpose. Here is yet another role for software development processes: while there is no
absolute guarantee, process quality is likely to lead to product quality, and the concern for
achieving and maintaining quality needs to permeate all software activities throughout the
software life-cycle.

Note: this is in line with Hall and Rapanotti’s view of assurance as a key factor in the
development process.

Related to software quality is a distinction which is often made between the functional and
non-functional characteristics that fit-for-purpose software should exhibit. Many such
characteristics have been identified and variously classified in the literature.

Note: ISO/IEC 9126-1:2001, an international standard for software product quality,

provides one such classification.

Some of these characteristics – those which are most relevant to this course – are listed
below.

● Functional fitness determines the extent to which a system does what the customer
wants and needs.
● Usability determines the effort required to learn about, operate, prepare input for
and understand the output of a system – that is, how easy the system is to use.
● Flexibility determines the effort required to modify an operational system – that is,
how easily the system can be changed while in service.
● Testability determines the effort required to test a system to ensure that it carries out
its intended function – that is, how easily the system can be tested to show that the
customer’s requirements have been met.
● Reusability determines the extent to which a system (or system component) can be
reused in other applications – that is, how easy it is to reuse some of the software to
make future developments more cost-effective.

Activity 10
Based on your own experience and practice, in what way do you think the various
activities forming part of the software development process may impact on the quality
of the released product?
Discussion
You may have covered some of the following points.
In analysis, validation of problem understanding and of what is required of the system
is key to quality, as this determines both the desired functionalities of the system and
its many non-functional characteristics. Eliciting knowledge directly from stakeholders,
e.g., users or businesspeople, and discussing our assumptions explicitly with them
would increase confidence in the accuracy of our understanding.
During design and implementation, following tried-and-tested approaches, good
practice and sound principles increases the chance of achieving high-quality software.
Access to experienced developers able to review the software is also a way to
maximise quality all round.

17 of 39 Wednesday 6 November 2019

5 Software quality

Comprehensive and systematic testing is essential to reduce the rate of occurrence of

software defects and generally verify overall software correctness. Various levels of
testing exist – some involving a wide range of stakeholders, including end users. In
some domains – for instance, safety-critical applications – the involvement of
independent certification authorities might also be required.
In maintenance, it is important to track and correct defects in the software which
become apparent after deployment, possibly providing end users with means to report
errors easily or make suggestions for product improvement.

Finally, while an appropriate investment in quality processes is important to software

quality, a quality culture is also required: an organisational environment in which the
concern for quality is highly valued and where there is support for developers who care for
all aspects of quality – even the less tangible ones such as code readability or adequacy
of documentation. A quality culture is characterised by highly professional behaviour
and ethical decision-making on the part of those involved in the process, and by the
allocation of an adequate level of resources for specific activities which contribute to
software quality: it is widely acknowledged that some software failures can be traced back
to excessive pressure from management to release a product which had not been
properly verified and validated, or to situations where developers were forced to cut
corners in the interest of quicker delivery. A quality culture implies an effective and mature
use of estimation, planning and risk management, and an integrated view of both human-
related and hard technological issues.

18 of 39 Wednesday 6 November 2019

6 Modelling and the UML

6 Modelling and the UML

‘Essentially, all models are wrong, but some are useful.’
(George Box)

Models are forms of description often adopted in software development. They are
abstractions used to represent and communicate what is important, devoid of
unnecessary detail, and to help developers deal with the complexity of the problem being
investigated or the solution being developed.
Modelling is used in other forms of design and engineering. For instance, architects
develop different models of buildings – some addressing structures, others materials,
others ergonomics, and so on. The same happens with modelling in software
development: some models are used to capture properties of the problem domain, such
as key elements of the business or how its processes work, while other models are used
to consider different aspects of the software, such as how the code is divided up and
organised, or how various elements of the software communicate and work together.
Each model is an abstract representation of some view of the system, and such views
may change as the development process unfolds.

Activity 11
Consider different types of model you have encountered in your own practice of
software development. What was their purpose? How were they expressed?

In software development, we build models from different perspectives. In particular, we

can distinguish between the following modelling types.
Domain modelling is concerned with understanding and modelling context information
for a specific problem, independently of a decision to use a software system to deal with
that problem. A domain model is a representation of the main concepts in the real-world
problem context – for instance, a business under consideration. A domain model does not
necessarily assume a software solution.
Specification modelling assumes that a software system will deal with the need in
context. A specification model represents software elements used in the software solution
to a problem, and is mainly concerned with the definition, at a high level of abstraction, of
the services provided by the software.
Design modelling describes the software system itself, with the allocation of
responsibilities to its various parts, and its behaviour and control flow.

Note: the distinction between ‘domain models’ and ‘design models’ was introduced by
Cook and Daniels (1994). They called the domain model the ‘essential model’, and the
design model the ‘implementation model’. ‘Business model’ is also often used in the
literature with a meaning similar to domain model.

Domain modelling plays an important role in understanding the need in context, before
suggesting any software solution. The important elements of the context, whether they
are people, products, departments, sensors, alarms or operators, need to be identified.

19 of 39 Wednesday 6 November 2019

6 Modelling and the UML

Only by understanding the need in context, and how, why and with what consequences
that need changes, can a fit-for-purpose software solution be specified. However,
modelling the domain may sometimes be unnecessary – for instance, when there is an
accepted need for a well-defined software system to solve a well-understood problem. In
such a case, modelling the domain would not bring much advantage.
Domain and specification modelling may produce very similar models, but the
interpretations of the models are different; the former types of model are about real-world
entities, the latter are about software representations of those entities.

Activity 12
How would you describe the relation between these three modelling perspectives and
the traditional split between analysis and design in software development?
Discussion
You may consider domain modelling and specification modelling as analysis activities
– that is, concerned with understanding what a problem is, and specifying what is
required of the software system to be developed. As you might expect, design
modelling corresponds to the design activity.

Models are usually constructed by following specific linguistic conventions, often referred
to as techniques, and the models’ level of formality will depend on the formality of those
conventions; for instance, models can be based on narrative, diagrams or even
mathematics.
Well known modelling techniques in software development are defined under theUnified
Modelling Language (UML). UML is one of the most popular and successful standards
currently used by the software industry. UML is the result of the merging of several
notations that appeared during the 1980s and early 1990s, augmented by new techniques
and even a mechanism to extend the notation further. You will not learn UML in its entirety,
but concentrate on a subset of the most commonly adopted techniques.

Note: to see the UML specification, click here.

It is important to note that UML is a modelling language, not a development process – it

gives you a set of techniques, but does not prescribe whether or how these techniques
should be used during development. In fact, because UML is the result of an exercise in
unification, it can be used flexibly within many different processes and practices, and the
same technique can serve different purposes. Also, UML models can be constructed at
different levels of precision. UML can be used equally effectively both as a sketching
notation, for example to jot down ideas or communicate them among stakeholders, and
for precise description of aspects of software systems, for example as semi-formal
specification of system functions.

Brief UML history

In the early 1990s, various object-oriented methods appeared along with a proliferation of
techniques and notations. By the mid-1990s two authors of well-known methods – James
Rumbaugh, one of the authors of the Object-Modelling Technique (OMT), and Grady
Booch, author of the Booch method (Rumbaugh et al., 1991; Booch, 1994) – joined forces

20 of 39 Wednesday 6 November 2019

6 Modelling and the UML

in an attempt to unify their methods. They named their joint effort the Unified Method.
Rumbaugh and Booch were soon joined by Ivar Jacobson and his colleagues, authors of
Object-Oriented Software Engineering (OOSE), in the development of what became known
as the Unified Modeling Language (UML); see Jacobson et al. (1992). As UML has evolved
it has incorporated feedback from the object community, and has won the support of many
people and organisations which considered the idea of a unified modelling language a
valuable one. UML was then submitted for standardisation to the Object Management
Group (OMG), a consortium of several large software companies that produces and
maintains computer-industry specifications. UML was formally adopted in 1997. The UML
specification is constantly updated. At the time of writing (2013), UML 2.5 is near
completion; this is the basis of the notation we use in this course.

21 of 39 Wednesday 6 November 2019

7 Object orientation

7 Object orientation
With the growth in code complexity came the idea of breaking up large sequential
programs into more manageable, and smaller, named pieces of code. Initially these
tended to represent actions, which in turn could be broken down into smaller and smaller
actions, until a unit was reached that was small enough to understand, write and test
independently. This type of code decomposition proved a useful device and became the
basis of procedural and functional styles of programming. Some of these are still in use,
although we will not consider them in this course.
The early1960s saw the idea of structuring software around the data structures or objects
that were manipulated, rather than around the actions that manipulated them. This was
the start of object-oriented programming, which subsequently developed into a set of
concepts, principles and techniques used throughout software development to analyse
software problems and design and construct software solutions. We use the term object
orientation to refer to this holistic approach.
You should note that because of its roots in programming, object orientation has a distinct
bias toward solution space descriptions. Nevertheless, over the years useful techniques
applicable across the problem/solution divide have been developed.
Although object orientation is relevant to many domains of application, it is seen as
particularly advantageous in business computing. It was observed that as businesses
change, the things their computer systems manipulate remain fairly stable, even though
the ways these systems are used may change rapidly. For instance, commercial systems
will most likely always need to represent taxes, goods, payments, prices and deliveries,
even though the business rules determining who is creditworthy, how orders are taken,
and whether payment is required before delivery may all change frequently. Therefore,
one advantage of structuring software around objects meaningful to the business is that it
gives more stability and better traceability from the business to the code.

7.1 Modelling with objects

In object orientation, modelling is very much focused on constructing descriptions based
on sets of interacting objects.
For instance, when exploring a real-world problem we identify the important objects which
relate to that problem and build models in terms of these objects and their interactions.
The purpose here is to get a better understanding of the domain, whether to elicit
requirements or to understand the business needs that those requirements address.
Ideally, the classes of object in our models should be chosen to correspond to natural
categories of things in the real world – such as ‘customer’, ‘invoice’, ‘payment’, ‘bill’ – so
that there will be a structural similarity between the real world and the software. This
approach can lead to good traceability from problem descriptions and requirements
through to code.
Later on, once we have assumed that a software system is needed to address the
problem, we may start modelling elements of the solution as software representations of
the real-world objects. Initially the emphasis is on the services provided by the software
solution and not on how these services are provided.

22 of 39 Wednesday 6 November 2019

7 Object orientation

As design progresses, modelling describes the software system in greater and greater
detail, with the allocation of responsibilities to different software objects so that the
functionality of the entire system is distributed among the classes. In object terms a
design model will represent the software classes and objects, the messages exchanged
between them and their order.
Note, however, that structuring the software around objects per se does not guarantee
quality software as an outcome. This can only be achieved through the judicious
application of sound design and construction principles, and indeed object orientation
comes equipped with a rich set of principles and tried-and-tested design solutions to
recurrent design problems.

7.2 Programming with objects

To represent real-world objects such as an account or a payment in software we construct
units of code, called objects, which assemble both data and operations that can act on
those data.

Note: this concept is in turn an evolution of the idea of programming with abstract data
types (Liskov and Zilles, 1974), which developed in the 1970s.

For instance, an object representing a bank account in software may include both a
variable holding the account’s current balance and a number of operations for credit and
debit transactions. Moreover, we expect that only those operations can access the data
directly: other software wanting to affect the current balance would have to make use of
those operations via an established interface. This particular feature of an object is called
encapsulation.

Activity 13
What do you think the advantages of encapsulation might be?
Discussion
Encapsulation provides an explicit boundary separating information about which
operations are available and information on how such operations are implemented. In
our example, outside the account object, available operations are advertised via their
interface only, which indicates how they can be invoked. The implementation of the
operations, i.e. how they materially access and modify data, and how the data are
represented through variables, are not visible outside the object. In this way, as long as
the interface remains unchanged, the operation implementation or the data
representation can be modified without affecting other software.

In most programming languages an object is defined as an instance of a homogeneous

class, with the class providing an abstraction for the common characteristics of all its
instances. For example, a class-defining current-account object might indicate that each
account will have a distinguishing account number and a balance, and that money can be
withdrawn or deposited; on the other hand, each of its instances will have its own distinct
account number and balance reflecting how much money is in the corresponding account.

23 of 39 Wednesday 6 November 2019

7 Object orientation

Activity 14
What do you think the advantages of distinguishing classes and objects might be?
Discussion
Classes allow for more succinct code and foster code reuse and flexibility. By defining
a class we are defining the properties and behaviour of all the objects of that class that
are created during software operation; when a change is needed, only the class code
needs to change, with such changes automatically propagated to all its instances.

No serious software system consists of a single object. Instead there will be a collection of
objects which collaborate to achieve the functionality of the whole system. Collaboration
occurs when one object requests a service from another object in order to perform some
task. To fulfil a particular collaboration, each object takes on a different role: the object that
makes the request can be seen as the client, and the object that receives the request
(and provides the service in response) as the supplier. The request is communicated to
the supplier from the client by message passing based on the objects’ interfaces: the
client sends a message to the supplier which on receipt executes a corresponding
operation; when the supplier has completed the execution of its operation it returns a
message answer to the client. Both the message and message answer may convey data
between the objects, with the interface specifying the type of the arguments and the result
of the operation.
Object collaboration through message passing is at the core of computation in an object-
oriented software system, and many forms of collaboration can be forged between
objects. In software development, a range of principles and techniques have been defined
to design and implement object collaborations resulting in quality software.

24 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

8 Finding and reading academic articles

The area of software development is fast-moving and as a software development
professional you cannot rely only on prepared materials to perform your role, but must
have strategies for finding materials that track the leading edge of your subject. Finding
relevant material is not a trivial task. There are myriad sources, including academic and
professional journals, conferences, technical reports, blogs, podcasts, newspaper and
magazine articles – the list is almost endless. Once your information resources have been
found, there’s still the investment of time needed to sift through the various resources,
which will be of varying relevance, so as to extract the best information in the least amount
of time. Then the task of understanding begins. Even once you’ve understood the
information contained in a resource (or think you’ve understood it), much of this
information’s value comes from relationships it will allow you to build between other
resources. Recording and developing those relationships – building your own literature –
is therefore also worthwhile.
In this last part of the course you will learn about a simple workflow for interacting with the
published academic literature on software development so as to underpin your
professional development and practice.

8.1 A workflow for reading the academic literature

The academic literature consists of papers and other resources, some of which you may
wish to read. To be read they need to be found. This section presents a simple workflow
by means of which the academic literature can be accessed.
There are five stages to this workflow:

Stage 1: Preparation
Stage 2: Discovery
Stage 3: Assimilation
Stage 4: Recording
Stage 5: Relating

These can be arranged as in Figure 1.

We will consider each stage in turn. As you work through this process you will find that the
assimilation stage itself consists of another workflow.

Preparation Discovery Assimilation Recording Relating

Figure 1 The M813 workflow for reading academic literature

The simple workflow outlined in this section can underpin research in many areas of
computing and in other science-based disciplines. The workflow is lightweight but
structures any future research you may do – for instance, as part of other taught or

25 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

research modules – and will underpin your studies as you progress through the
programme.

8.1.1 Preparation
There are many tools available for creating and managing bibliographies. Indeed, you
may already use a bibliographic database management system (BDMS) such as Zotero,
EndNote, BibDesk or RefWorks. If you already use a BDMS then you are free to use it in
this part of the course.
BDMSs are complex tools that can hold thousands of academic papers, white papers and
many other information-resource items, keeping them for easy reference. As well as
storing resources, some BDMSs allow an electronic version of an information resource to
be kept alongside notes you have made about that resource. The resource and any
associated notes are then ready for use when you need to cite the material. In some
document preparation systems (e.g. Microsoft Word) this can be done automatically.
Even the simplest BDMS requires great flexibility and is a very complex piece of software.
The learning curve can be quite steep. For this reason, if you don’t already use a BDMS
we suggest that you create a simple word-processing document to store the details of the
few papers you might find useful to go alongside this course - perhaps something from the
references that interested you.
If you already use a BDMS, you can create a new database for this course. Alternatively,
you could simply copy and paste the template below into any Word/text-processing
package of your choice for each new information resource that you keep in your
bibliography.

CiteKey:
Title:
Author(s):
Format:
Publisher:
Year:
URL and last accessed:
Referenced by:
Abstract:
Notes:

Activity 15
What do you think the fields in the template capture?
Discussion
The template consists of a single table with the following fields (this structure is a
subset of any commercial tool’s structure, although in such a tool there may be many
other fields too):

● CiteKey: a unique identifier (or primary key) for the entry

● Title: the title of the work

26 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

● Author(s): the author(s) of the work, presented if necessary as a list

● Format: the type of resource, i.e. whether a journal, magazine, technical report,
commercial white paper, web page, personal communication, etc.
● Publisher: the publisher of the resource
● Year: the year the resource was published or the year it was last checked
● Pages: the page numbers, if applicable
● URL and last accessed: the web location where the resource was found, and the
date on which you last accessed it
● Referenced by: which other item(s) in your BDMS (if any) suggested this
information resource
● Abstract: a copy of the abstract for the resource, should one exist
● Notes: the notes that you have made on the resource.

8.1.2 Discovery
Discovering relevant academic literature is becoming easier and easier as most
bibliographical collections are now accessible online. Most libraries have subscriptions
which allows users to access copyright published materials: if you are currently studying
an OU module, you should explore the rich catalogue of sources and resources which are
available to you through the OU Library. Search engines also exist which specialise in
searching academic literature online. Once such engine is Google Scholar, which you will
use in the next activity.

Activity 16
Use Google Scholar from your web browser to search for Srinivasan Keshav’s article
titled ‘How to read a paper’: this is an excellent, easy-to-read introduction to an
effective workflow for reading many research articles in the field of science and
management (as opposed to, for instance, arts). Once you have found it, record as
much information as you can in your newly created BDMS: you’ll enhance this record
later on, in the fourth step of your workflow, after engaging with the material.
Discussion
Hopefully you managed to find the paper without too much difficulty. The paper is
included in a print journal and its full bibliographical reference is:

Keshav, S. (2007) ‘How to read a paper’, ACM SIGCOMM Computer

Communication Review, vol. 37, no. 3, pp. 83–4.

This can be recorded using our template as:

CiteKey: Keshav2007How
Title: How to read a paper
Author(s): S. Keshav ([email protected])
Format: Academic paper
Publisher: ACM SIGCOMM Computer Communication Review
Year: 2007
Pages: 83-84

27 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

This published version is copyright materials and its text can only be accessed if you
are a licensed user, for instance an OU student using the OU Library. For this course,
however, we have obtained permission from the author to reproduce the content of the
article, which you can find in the Appendix.

Note: that, in general, Google Scholar will not necessarily find everything that could be
available, as some publishers do not let Google’s crawlers see their content. However,
more significantly, Google Scholar may actually return a preprint or draft version of an
article rather than the final published piece. Books are sometimes retrieved as a digital
part-version rather than the full text.

8.1.3 Assimilating an academic article

Some of the most challenging information resources that exist come from the academic
literature: even the most experienced researcher will not expect to read an academic
paper once and immediately understand it. Accordingly, the workflow for assimilating the
information is complex. In this subsection you will explore this complexity by using a
workflow suggested by the author of the paper you have just ‘discovered’.
The barriers to understanding a paper are many and high. An academic paper is
considered good if it makes a contribution to knowledge; therefore, to understand an
academic paper is to understand the contribution to knowledge that it makes. This may
require a substantial investment of time and be very challenging.
Understanding an academic (or other) article may require you to understand a new
concept or new terminology, to look at an existing model, concept or ‘thing’ from a new
perspective, or to see a group of things in a new relationship. Accomplishing tasks such
as these will very probably take more than one reading.
If you were on holiday – for instance – you might be able to dedicate a full weekend or a
week to an academic article and read it through as many times as you need. However,
you will probably not always be able to dedicate a block of time to the many readings an
individual academic paper might need, and so you must find a way of recording your
understanding of a paper from one reading to the next.

Annotation
Here are the annotations we made on the published version of Keshav’s paper. They
appear in red and as electronic highlights in yellow.

28 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

Figure 2 Our annotation of Keshav’s article (Keshav, 2007)

Activity 17
By looking at the text of the notes, identify and list the three dates on which the paper
was read. What do you think we were trying to convey with the annotation?

Provide your answer...

Discussion
The dates are:

● First read: 21 December 2012

● Second read: 23 December 2012
● Third read: 15 January 2013

29 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

The notes appear to record observations on the workflow, and any differences from
what the reader had expected. For instance, in two cases an initial note is followed up
with a further note. The first case has:

● ‘23/12/12: A good workflow.’

followed up by:

● ‘15/01/13: we say it may take many, i.e., more that [sic] three, passes.’

The second note has:

● ‘23/12/12: a first pass may not be enough to know this.’

followed up by:

● ‘15/01/13: ok, of relevance someday.’

The first pairing of notes suggests that in the reader’s experience, you sometimes
have to read an academic paper several times before you really understand it.
The second pairing of notes shows the reader’s growing understanding of what the
purpose of the first pass is.

Keshav (2007) advocates following a workflow for reading an academic paper. There are
three passes to this workflow, and the subsections that follow will examine these as a way
of assimilating the information presented.

8.1.4 Keshav’s first pass

Keshav writes (emphasis added):

The first pass is a quick scan to get a bird’s-eye view of the paper. You can also
decide whether you need to do any more passes. This pass should take about
five to ten minutes and consists of the following steps:

1. Carefully read the title, abstract, and introduction

2. Read the section and sub-section headings, but ignore everything else
3. Read the conclusions
4. Glance over the references, mentally ticking off the ones you’ve already
read.

Keshav adds that at the end of the first pass, the reader should be able to answer ‘the five
Cs’, which are:

● Category: What type of paper is this? A measurement paper? An analysis of an

existing system? A description of a research prototype?
● Context: Which other papers is it related to? Which theoretical bases were used to
analyse the problem?
● Correctness: Do the assumptions appear to be valid?
● Contributions: What are the paper’s main contributions?
● Clarity: Is the paper well written?

30 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

Activity 18
Apply the first stage of the workflow to Keshav’s own paper: carry out a first pass and
see if you can answer the five Cs, i.e., Category, Context, Correctness, Contributions
and Clarity.

Provide your answer...

Discussion
We came up with the following. The five Cs are:
Category: Keshav’s paper analyses a process that researchers can use to read an
academic article. The paper is an experience report of the reader, in consultation with
some academic colleagues. It provides a contribution to practice.
Context: the paper does not include a bibliography of academic papers; rather, there is
a list of website resources that relate the paper to other contributions to practice in this
area.
Correctness: assumptions stated in the abstract are:

● that researchers spend a great deal of time reading research papers

● that this skill is rarely taught, which leads to wasted effort.

Contributions: the paper’s main contribution is that it outlines a practical and efficient
three-pass method for reading research papers. There is a subsidiary contribution: it
describes how to use the three-pass method to undertake a literature survey.
Clarity: as is evident from the headings used, the paper’s structure is generally logical.
However, there is no conclusion, so we struggled to know what to do to complete the
first pass.

At this point in the process, given the five Cs, you should be able to decide whether
whatever paper you are reading is interesting and relevant. As for Keshav’s paper, since
we will be using it for the remainder of this section, we can answer the question as to
where it is interesting or relevant with a resounding ‘yes’ (whether you actually find the
paper interesting is irrelevant at this point).

8.1.5 Keshav’s second pass

Keshav says that the second pass should take about an hour. The goal of the second
pass is to be able to summarise the ‘main thrust of the paper, with supporting evidence, to
someone else’, i.e., to be able to reconstruct the reasoning that the paper uses in its
contribution to knowledge or practice.
Keshav advises the reader to take greater care on the second pass, but to ignore highly
technical details such as proofs, implementations or detailed justification of any argument
– indeed, anything that will extend outside of the one-hour window you have to read the
paper a second time.
Keshav proposes indicators of a paper’s quality. For him, figures, diagrams and other
illustrations should have been carefully set out so as to give the reader all the information
he or she needs.

31 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

Finally, Keshav suggests marking interesting references, so that the portion of the
literature upon which the arguments of the paper rest can be read (see Figure 3).

23/12/12:
2/12:
Interesting
resting

Figure 3 The references list at the end of Keshav’s paper (Keshav, 2007)

Activity 19
Do a second pass through Keshav’s paper. Try to explain your understanding of the
paper to someone else.
Discussion
In trying to explain Keshav’s work, we came up with the workflow in Figure 4 (here’s a
scalable PDF download of this image).

Figure 4 Our diagrammatic representation of Keshav’s workflow

This paper’s workflow depends upon various conventions being followed by an
academic article and which are to all intents and purposes followed in the sciences
(including software development as an academic discipline). In other areas of software
development such conventions may not necessarily hold – for example, commercial
(or white) papers may deviate.
The conventions assumed of a written academic item are that it will have:

● An abstract, which is a brief summary of the item – whether it be a research

article, thesis, review, conference proceeding or any in-depth analysis of a
particular subject or discipline. The abstract is used to help the reader quickly
ascertain the item’s purpose.

32 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

● A number of sections, beginning with an introduction and ending with conclusions

and/or a discussion.
● A bibliography, essentially a list of literature and other material, such as web
pages (and even personal communications between researchers) to which the
reader of the article can refer to check the arguments and evidence that others
have presented and which form the paper’s basis.

8.1.6 Keshav’s third pass

It may be that, for Keshav’s paper at least, you’ve already gained enough understanding
to make a third pass unnecessary. However, for most academic papers this won’t be true
– moreover, for some of the most complex research articles a third, or even fourth, fifth or
sixth pass still will not suffice. Academics have been known to spend their whole working
lives trying to understand (and build on) the research covered by a single paper.
Keshav offers some very useful observations:

Sometimes you won’t understand a paper even at the end of the second pass.
This may be because the subject matter is new to you, with unfamiliar
terminology and acronyms. Or the authors may use a proof or experimental
technique that you don’t understand, so that the bulk of the paper is
incomprehensible. The paper may be poorly written with unsubstantiated
assertions and numerous forward references. Or it could just be that it’s late at
night and you’re tired.
You can now choose to: (a) set the paper aside, hoping you don’t need to
understand the material to be successful in your career, (b) return to the paper
later, perhaps after reading background material or (c) persevere and go on to
the third pass.

Decision criteria for a third pass

As Keshav argues, not all papers require a third pass and you should take care in
deciding whether a particular paper should have one at this particular point. We say this
because the importance of understanding a paper may only become apparent later – it
may turn out to be a highly cited paper that needs to be understood so that other papers
become accessible; or, it may be that a career change on your part means you need to
come back to it. Here we see the motivation for annotating a paper and recording your
present understanding: if you find you do need to return to the paper at some point, you
won’t start from scratch if you’ve made good notes. This also makes clear why you should
keep a master version of a paper, with your earlier notes stored with it.
If you decide that investing time in a third pass of a paper is indeed necessary, Keshav’s
key skill is to attempt to ‘virtually re-implement the paper’ – that is, making the same
assumptions as the authors: recreate their work. This means being able to restate their
assumptions, and to recreate the reasoning they use to reach their conclusions.
Essentially, you are asking whether, starting at the same point, you could have made the
same contribution to knowledge or practice as the authors. As an aside, that’s a good
definition of truly understanding an academic paper.

33 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

There are many reasons why you might not have been able to make the same
contribution, by the way – and not all of them point to a lack of understanding (or creative
genius) on your part:

● perhaps the paper has failings, makes inappropriate assumptions, and/or fails to cite
other work that you know of
● perhaps the arguments the paper puts forward are circular and/or nonsensical
● perhaps the research method used does not make sense.

Keshav adds: ‘During this pass, you should also jot down ideas for future work’. This is an
excellent idea, should you be a researcher (as Keshav is), or a practitioner, and you may
wish to identify projects or the like to which the understanding you have gained can be
applied.
According to Keshav, the third pass will take about four or five hours for a beginner, and
about an hour for an experienced reader. However, longer periods are quite possible:
indeed, whether even more passes are required will very much depend on how critical to
you the paper is.

Activity 20
Carry out a third pass on Keshav’s paper, and then try to construct a virtual re-
implementation of the paper.

8.1.7 Recording
Recording your understanding of an academic paper is an important aspect of reading it.
Being able to relate it to the literature you already know – that is, contextualising the paper
in the literature – is another. At the moment, of course, your knowledge of the academic
(and other) literature may be very small – or indeed non-existent. Everyone has to start
somewhere. Here you will begin recording your understanding of the academic literature
in your BDMS.

Activity 21
Having read and annotated Keshav’s paper, update the entry in your BDMS with your
notes.
Discussion
Having added the abstract and our notes, our own entry now reads:

CiteKey: Keshav2007How
Title: How to read a paper
Author(s): S. Keshav ([email protected])
Format: Academic paper
Publisher: ACM SIGCOMM Computer Communication Review
Year: 2007
Pages: 83–84

34 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

URL and last accessed:

http://dl.acm.org.libezproxy.open.ac.uk/citation.cfm?doid=1273445.1273458 Last
accessed: 19 November 2013.
Referenced by: OpenLearn free course materials
Abstract: ‘Researchers spend a great deal of time reading research papers.
However, this skill is rarely taught, leading to much wasted effort. This article outlines
a practical and efficient three-pass method for reading research papers. I also
describe how to use this method to do a literature survey.’
Notes: An excellent resource that provides an efficient workflow for reading an
academic article.

As you go on collecting information resources, you’ll find that it is worth being critical
about what you retain in your BDMS – not all information resources are worth keeping. It
is also good to be systematic. It is good practice to record a resource directly after having
found it; then record your thoughts after the first pass and update your notes after
subsequent passes.
Your BDMS needs to both discriminate and retain a level of understanding of the
resource. With this in mind, we have positioned the Recording stage after the Assimilation
stage (although, as noted, you may make an initial note of the discovery at an earlier
point).

8.1.8 Relating
As you will realise from your professional practice, there are many contributors to any
area of research or practice. The relationship between contributors is complex
(sometimes circular), extends in time and can be affected by politics. At this point,
however, we will look only at indicators of dependence as revealed through a bibliography.
The bibliography of a paper indicates the parts of the context of the paper that have been
influential in its writing. Keshav’s bibliography consists of four websites. Such a source
citation is appropriate for his paper, but actually is not typical of an academic paper. Even
so, we can plot the relationship between Keshav’s paper and the material it cites as in
Figure 5.

Keshav

ACM SIGCOMM
Roscoe Schulzrinne Whitesides Computer
Communication
Review Online

Figure 5 The relationship between Keshav’s paper and the material it references

35 of 39 Wednesday 6 November 2019

8 Finding and reading academic articles

Activity 22
Add each element in Keshav’s bibliography to your BDMS.
Discussion
You may have noticed that the Roscoe website itself lists as a resource the Keshav
paper that you have just analysed. That means that if we were to extend Figure 5 there
would be a loop in it. You would not often find this phenomenon in typical archival
research articles because a paper is normally printed and once printed is unchange-
able. Online resources such as Roscoe’s can of course be updated, so can
incorporate references to other resources that were created after the original item.
We added five entries to our BDMS, corresponding to the five references. For each we
set the ‘Referenced by’ field to ‘Keshav2007How’ as suggested by the diagram we
drew.

Now complete Activity 23, which returns you to the discovery stage.

Activity 23
Consider a few topics that interest you in the area of software development. Focus on
one of them, and using a search method of your choice, find three research articles
that are relevant to your interests. Run through Keshav’s workflow (first pass only) for
each item found and if you think the item is sufficiently interesting, create an entry in
your BDMS. Record your notes on each.
Do you think that you’d conduct a second (or third) pass on any paper that you found?

36 of 39 Wednesday 6 November 2019

Conclusion

Conclusion
In this course we have looked at some key concepts, themes and skills related to software
development, as an adapted extract from the Open University course M813 Software
development.
Should you decide to study further, this course will allow you to develop the fundamental
knowledge, understanding, and analysis and synthesis skills that you need to develop fit-
for-purpose software in an organisational context, by taking a practice-based approach
based on an organisation you are familiar with. The course will also give you an
opportunity to investigate emerging trends in software development and carry out some
independent research into issues in software development that interest you.

References
Bass, L., Clements, P. and Kazman, R. (2003) Software Architecture in Practice, Addison-
Wesley.
Beck, K. (2000) Extreme Programming Explained, Addison-Wesley.
Benington, H. D. (1983) ‘Production of large computer programs’, Annals of the History of
Computing, vol. 5, no. 4, pp. 350–61.
Bird, P. (1994) LEO: The First Business Computer, Workingham, UK, Hasler Publishing.
Boehm, B. W. (1988) ‘A spiral model of software development and enhancement’,
Computer, vol. 21, no. 5, pp. 61–72.
Booch, G. (1994) Object-Oriented Analysis and Design, Addison-Wesley.
Cockburn, A. and Highsmith, J. (2001) ‘Agile software development: the people factor’,
Computer, vol. 34, no. 11, pp. 131–133.
Cook, S. and Daniels, J. (1994) Designing Object Systems: Object-Oriented Modeling
with Syntropy, Upper Saddle River, NJ, Prentice-Hall.
Forsberg, K. and Mooz, H. (1992) ‘The relationship of system engineering to the project
cycle’, Engineering Management Journal, vol. 4, no. 3, pp. 36–43.
Fowler, M. and Highsmith, J. (2001) ‘The Agile manifesto’, Software Development, vol. 9,
no. 8, pp. 28–35.
Grant, R. M. (2007) Cases to Accompany Contemporary Strategy Analysis, 6th edn,
Wiley.
Gunter, C. A., Gunter, E. L., Jackson, M. and Zave, P. (2000) ‘A reference model for
requirements and specifications’, Software, IEEE, vol. 17, no. 3, pp. 37–43.
Hall, J. G. and Rapanotti, L. (2003) ‘A reference model for requirements engineering’, in
Proceedings of the 11th International Requirements Engineering Conference, 2003.
Montery Bay, CA, September 8–12 2003. New York, IEEE, pp. 181–7.
Hall, J. G. and Rapanotti, L. (2009) ‘Assurance-driven design in problem oriented
engineering’, International Journal on Advances in Systems and Measurements, vol. 2,
no. 1, pp. 119–30.
Jackson, M. (1995) ‘The world and the machine’, in ICSE 1995: 17th International
Conference on Software Engineering. Seattle, WA, April 23–30 1995. New York, IEEE,
pp. 283–92.

37 of 39 Wednesday 6 November 2019

Acknowledgements

Jacobson, I., Booch, G. and Rumbaugh, J. (1999) The Unified Software Development
Process, Addison-Wesley.
Jacobson, I., Christerson, M., Jonsson, P. and Overgaard, G. (1992) Object-Oriented
Software Engineering: A Use Case Driven Approach, Addison-Wesley.
Keshav, S. (2007) ‘How to read a paper’, ACM SIGCOMM Computer Communication
Review, vol. 37, no. 3 [Online]. Available at
http://dl.acm.org.libezproxy.open.ac.uk/citation.cfm?doid=1273445.1273458 (Accessed
19 November 2013).
Liskov, B. and Zilles, S. (1974) ‘Programming with abstract data types’, ACM Sigplan
Notices , vol. 9, no. 4, pp. 50–9. ACM.
McManus, J. and Wood-Harper, T. (2008) A Study in Project Failure: Technical
Report, BCS.
Robertson, S. and Robertson, J.C., (2006) Mastering the Requirements Process, 2nd
edn, Addison-Wesley.
Rogers, G. F. C. (1983) The Nature of Engineering: A Philosophy of Technology, Palgrave
Macmillan.
Royce, W. (1970) ‘Managing the development of large software systems’, Proceedings of
IEEE WESCON, vol. 26, pp. 1–9.
Rumbaugh, J., Blaha, M., Premerlani, W., Eddy, F. and Lorensen, W. (1991) Object-
Oriented Modeling and Design, Prentice Hall.
Schwaber, K. and Beedle, M. (2001) Agile Software Development with Scrum, Prentice
Hall.
Selic, B. (2003) ‘The pragmatics of model-driven development’, Software, vol. 20, no. 5,
pp. 19–25.
Turski, W. (1986) ‘And no philosopher’s stone either’, Information Processing ’86:
Proceedings of the IFIP Congress, Dublin, Sept. 1–5, 1986, London, Elsevier, pp. 1077-
–1080.
Online only
Enticknap, N. (1998) ‘Computing’s Golden Jubilee’, Resurrection, no. 20 [online].
Available at http://www.cs.man.ac.uk/CCS/res/res20.htm (Accessed 8 July 2014).
Lake, M. (2010) ‘Epic failures: 11 infamous software bugs’, ComputerWorld [online].
Available at www.computerworld.com/s/article/9183580/Epic_failures_11_infamous_soft-
ware_bugs (Accessed 21 May 2013).
Standish Group (2009) CHAOS Summary 2009 [online]. Available at http://blog.
standishgroup.com (Accessed 25 March 2013).
Tootill, G. (1998) ‘The original original program’, Resurrection, no. 20 [online]. Available at
http://www.cs.man.ac.uk/CCS/res/res20.htm (Accessed 8 July 2014).

Acknowledgements
This course was written by Lucia Rapanotti and Jon G. Hall.
Except for third party materials and otherwise stated (see terms and conditions), this
content is made available under a
Creative Commons Attribution-NonCommercial-ShareAlike 2.0 Licence.

38 of 39 Wednesday 6 November 2019

Acknowledgements

Course image: Texture X in Flickr made available under

Creative Commons Attribution 2.0 Licence.
The material acknowledged below is Proprietary and used under licence (not subject to
Creative Commons Licence). Grateful acknowledgement is made to the following sources
for permission to reproduce material in this course:
Figure 2: Keshav, S. (2007) ‘How to read a paper’, ACM SIGCOMM Computer
Communication Review, vol. 37, no. 3 [Online], ACM New York, ACM DL Digital Library.
Figure 3: Keshav, S. (2007) ‘How to read a paper’, ACM SIGCOMM Computer
Communication Review, vol. 37, no. 3 [Online], ACM New York, ACM DL Digital Library.
Every effort has been made to contact copyright owners. If any have been inadvertently
overlooked, the publishers will be pleased to make the necessary arrangements at the
first opportunity.
Don't miss out:
If reading this text has inspired you to learn more, you may be interested in joining the
millions of people who discover our free learning resources and qualifications by visiting
The Open University - www.open.edu/openlearn/free-courses

39 of 39 Wednesday 6 November 2019