Abstract As code is now an inextricable part of science it should be supported by competent Software Engineering, analogously to statistical claims being properly supported by competent statistics. If and when code avoids adequate scrutiny, science becomes unreliable and unverifiable because results — text, data, graphs, images, etc — depend on untrustworthy code. Currently, scientists rarely assure the quality of the code they rely on, and rarely make it accessible for scrutiny. Even when available, scientists rarely provide adequate documentation to understand or use it reliably. This paper proposes and justifies ways to improve science using code: 1. Professional Software Engineers can help, particularly in critical fields such as public health, climate change and energy. 2. ‘Software Engineering Boards,’ analogous to Ethics or Institutional Review Boards, should be instigated and used. 3. The Reproducible Analytic Pipeline (RAP) methodology can be generalized to cover code and Software Engineering methodologies, in a generalization this paper introduces called RAP+. RAP+ (or comparable interventions) could be supported and or even required in journal, conference and funding body policies. The paper’s Supplemental Material provides a summary of Software Engineering best practice relevant to scientific research, including further suggestions for RAP+ workflows. ‘Science is what we understand well enough to explain to a computer.’ Donald E. Knuth in $A=B$ [ 1] ‘I have to write to discover what I am doing.’ Flannery O’Connor, quoted in Write for your life [ 2] ‘Criticism is the mother of methodology.’ Robert P. Abelson in Statistics as Principled Argument [ 3] ‘From its earliest times, science has operated by being open and transparent about methods and evidence, regardless of which technology has been in vogue.’ Editorial in Nature [4]

A national example of open source digital healthcare is critiqued. The code for implementing numeric patient identifiers is surprisingly naïve and bug-ridden, despite patient identifiers being computationally trivial and a critical component of reliable healthcare. The issues raised are shown to be widespread, long term, and apparently unrecognized. Problems are traced back to inadequacies in the relevant standards, and, at every stage, regulation through to development, inadequate Software Engineering input. An important finding is that the relevant healthcare standards are inconsistent and written without sufficient rigor to be at all constructive for implementing digital systems. The widely recognized problems of interoperability may be traced back to diverse (and buggy) interpretations of vague standards.

The study of Human Factors (HF) and Interaction Design (ID) plays a central role in critical systems design. HF discovers and applies information about human behaviour, abilities, limitations and other characteristics to the design of tools, machines, systems, tasks, jobs and environments, for productive, safe, comfortable and effective human use. Successful ID is inherently multidisciplinary, forward looking, and aims to sketch, synthesise and prototype the future. Although the two fields are closely related, there are critical differences in approaches. Interaction designers typically seek to shape, create and explore future solutions, whereas HF researchers seek to operationalize social, psychological and behavioural theory to optimize design, often with constraints, such as error-free interaction, generally for skilled workers. HF work has a long tradition in the workplace, often concerned with dangerous and critical activities performed by skilled operators, whereas ID is increasingly focused on the huge market of discretionary consumers, often concerned with the likes and dislikes of people.

Number entry is a mundane and error-prone task. To find errors, users often rely on visual checks to compare the differences between their instructions and the numbers they have actually input, a task that is difficult for users to do accurately. We therefore propose the use of number-based graphical representations (GRs) as a complement to conventional numeric representations (NR) to enhance visual checks, so users can examine both GRs and NRs to detect errors. We conducted two experiments to explore the issues raised. Experiment 1 examined the effects of GRs and NRs on representation difference detection (i.e. checking if two GRs or NRs are identical). The two representations had a comparative performance by time and error rate. In Experiment 2, we investigated the performance of GRs and NRs with number entry tasks. While extending the task time (increased by 38%), number entry with GRs resulted in significantly fewer errors than without GRs (decreased by 60%). Participants also had a high preference for number entry with GRs. Therefore, the proposed technique is promising for number entry error reduction, and that in safety critical applications improved safety can be achieved.

Managing and reducing error in number entry tasks is important, especially in safety critical contexts. Understanding factors that can affect number entry accuracy could help the design of more dependable systems. We present three interventions for number entry tasks, inspired by known priming effects (where exposure to prior stimuli can impact task performance). The interventions were questions for the operator to answer before entering each number. Questions related to the value/size of the number, its structure, and the context of the number entry task respectively. Results of a within-subject study show that although there was no significant difference amongst performance across interventions, all three interventions helped to improve the accuracy of number entry by reducing entry errors (by up to 40.8%) and unnoticed errors (by up to 60.7%). These are impressive gains and suggest the importance of more work in this area.

The world is faced with a growing number of people who live with chronic medical conditions. There have been numerous digital interventions into personal management of these diseases in recent years, yet gaps remain in the HCI literature. In particular, we lack a systematic understanding of user requirements in tools that support independent management while away from external influences. This paper presents a first investigation into low-intervention support for self-management. A mobile application enabled individuals to capture contextual information related to their health in the form of photographs. Through a month-long user study, we identify four management trends amongst our participants and describe their influence on mobile application adoption.

Numbers are used in critical applications, including finance, healthcare, aviation, and of course in every aspect of computing. User interfaces for number entry in many devices (calculators, spreadsheets, infusion pumps, mobile phones, etc.) have bugs and design defects that induce unnecessary use errors that compromise their dependability. Focusing on Arabic key interfaces, which use digit keys 0-9-· usually augmented with correction keys, this paper introduces a method for formalising and managing design problems. Since number entry and devices such as calculators have been the subject of extensive user interface research since at least the 1980s, the diverse design defects uncovered imply that user evaluation methodologies are insufficient for critical applications. Likewise, formal methods are not being applied effectively. User interfaces are not trivial and more attention should be paid to their correct design and implementation. The paper includes many recommendations for designing safer number entry user interfaces.

Seven segment number displays are ubiquitous and popular. They are simple and familiar. They seem to make economic sense, and with only seven segments they require little wiring and electronics to support. They are cheap to buy and cheap to use; they make seemingly effective and unproblematic products.

Although number entry appears to be a trivial task, user errors are still common and could be a result of poorly engineered interaction with the devices. We are challenging the design of universal number entry systems by looking at cases where user errors are frequently made. The telephone is used as a platform to compare input devices for number entry where we can look for speed and accuracy trade-offs between direct and indirect inputs. We will focus on the knob, button, and touchscreen and hope to find guidelines for when each is appropriate to use in a number entry system.

Mobile reading is becoming evermore popular with the introduction of eInk devices such as the Kindle, as well as the many reading applications available on slate PCs and cellular handsets. The portable nature and large storage capacity of these modern mobile devices is making reading a more technology orientated activity. One aspect of mobile reading that has been given surprisingly little attention is collective reading - which is a common activity with paper documents. We investigate the support of group reading using slate PCs, focussing on collective annotation. In the past, desktop PCs have proved inferior in many ways for reading, when compared to paper. Notably, user evaluations of our new system, BuddyBooks, demonstrate that the slate PC form factor can, in contrast, provide advantages for group reading. While annotation practices change with the new format, coordinating within the group can be improved when touch-interaction is carefully exploited.

Attentive reading is a complex and cognitively demanding task that uses note-taking and annotation to support the reader’s interpretation of the document. When reading on paper, extensive use of highlighting and other activities are conducted to support attentive reading, but this rich behaviour is not used with digital documents. Many users therefore print digital documents and then interact with them in physical form. This paper presents the “Digital Reading Desk,” an enhanced digital reading environment that provides support for attentive reading, providing a large working space for notes that mimics the use of desk space in a conventional physical environment. The Reading Desk uses a single tool to support both annotation and bookmarking, simplifying both the user’s learning of the system and their use of tools. Evaluation of the Reading Desk indicates preference for it, as well as considerable behavioural differences between our approach and traditional digital reading tools.

SUMMARY Software should be correct and robust. This paper suggests that we need forthright words for the failure of not being robust—heedless and heedlessness—and of recursively creating software (such as a compiler or virtual machine) that itself does not support dependable software development. Heedless programming is common, particularly affecting ‘trivial’ operations such as on numbers, and extends deep into programming language design and into the use of computers more widely, thus making robust, dependable applications of all sorts unnecessarily problematic. The paper defines the problem and presents a call to action to start addressing the problems identified. Copyright © 2012 John Wiley & Sons, Ltd.

Reading is increasingly being performed interactively on-screen; for instance, new novels are now routinely released in electronic format for viewing on PCs and mobile devices. Unfortunately, on-screen reading loses many of the natural features of conventional physical media, such as the ability to annotate, slip in bookmarks, turn page corners, and so on. How best should these features be represented electronically? Can computerized representations give benefits that excel the conventional benefits of paper? We describe the design and implementation of a novel reading system that mimics key properties of paper and surpasses them by incorporating digital techniques. A comparative user study evaluating the system confirmed the effectiveness of the features and the value of the system as a whole.

Signposts make complex documents more flexible and easier to read; signposts exist in many forms, but are familiar as cross references and as hypertext links. Signposting systems need to be carefully designed, so that they are reliable, easy to use for readers yet convenient for authors to embed in their writing. There are fundamental problems in achieving good signposting, which this paper explores through cross-referencing in the popular medium of LaTeX, which is often the typesetting system of choice for heavily cross-referenced documents. This paper provides an implementation of cross references for LaTeX users. Problems with LaTeX and TeX arise, and are explored and mostly solved: some problems are due to the designs of these systems, but, crucially, some problems are unavoidable, an inevitable part of signposting in principle, whatever systems are used, even including the web.

Motivation– Designers make decisions that later influence how users work with the systems that they have designed. When errors occur in use, it is tempting to focus on the active errors rather than on the latent design decisions that framed the context of error, and fixing latent conditions can have a more general (and future) impact than addressing particular active failures. Research approach– A constructive computer science approach is used, and results from a simulation reported. Research limitations– Error is a complex multidisciplinary field; this article makes a new contribution complimentary to human factors engineering. Take away message– This article shows that latent design decisions cause serious problems (including fatalities) in safety critical applications; the article proposes UI discovery tools to identify and manage latent errors. UI discovery enables human factors engineers and programmers to work together to help eliminate broad classes of latent design errors.

The field of inclusive design relates the capabilities of the population to the design of products by better characterising the user–product relationship. Inclusion refers to the quantitative relationship between the demand made by design features and the capability ranges of users who may be excluded from use of the product because of those features. By 2020, almost half the adult population in the UK will be over 50, with the over 80s being the most rapidly growing sector. These “inclusive” populations contain a great variation in sensory, cognitive and physical user capabilities, particularly when non-age-related impairments are taken into account. While this sector of the user population is growing, the most productive research and design sector is inevitably shrinking in proportion to that growth. In user interface design we develop interactive products and we are used to designing “for the user,” but often the “user” is rarely precisely specified demographically, and rarely described in a way that represents the variance in abilities. At worst, we may think that the users are a bit like us, but maybe do not know so much; and the environment where designers work is very different from the environment where users live. Users are often under contextual pressures that designers cannot relate to except through proper usability studies. Consider entering a hard-to-remember numerical code into a noisy burglar alarm to silence it in the 10 s allowed, working in a dark cupboard with a small hard-to-read LCD display. Unfortunately, most usability studies in the literature come from university laboratories and use students as experimental participants, which is convenient but ignores the fact that almost all are under 30! The problem in interaction design mirrors the recent observation that painkiller research tends to be done on young men and almost never on pregnant women; thus painkillers are developed, evaluated and refined on population sectors that are not representative of all users, and hence for perfectly understandable reasons they are becoming less suitable for some sectors of the population. But one needs to quantify these issues carefully and find suitable ways to address the identified problems.

Affordance is a widely used term in human–computer interaction (HCI) that, while familiar and attractive, does not have a clear operational definition. Using the mathematical concept of symmetry, this paper shows that it is possible to begin developing an operational definition for significant aspects of affordance by forming the theoretical concept of symmetry-affordance. The proposed definition restricts symmetry-affordance to particular contexts but in doing so makes it more useful, as it is clear how to exploit symmetry to aid design. The definition is in standard mathematics (in fact, group theory and model theory) and requires little additional structure. In examining symmetry-affordance, it becomes clear that some other HCI notions can be similarly interpreted by symmetry. The paper provides examples and design insights. Symmetry, as wide or as narrow as you may define its meaning, is one idea by which man through the ages has tried to comprehend and create order, beauty, and perfection. Hermann Weyl [(1952) Symmetry, Princeton University Press, Princeton, NJ]

Science fiction stories seductively portray robots as human. In present reality (early 21st century) robots are machines, even though they can do many things far better than humans (fly, swim, play chess to name a few). Any ethics for or of robots is therefore a seductive mix of fiction and reality. The key issue for rational discourse is to provide a rigorous framework for reasoning about the issues, including identifying flaws in the framework. We find such meta-reasoning in discussion about robot ethics to be ready for improvement. This paper takes its inspiration from B. Whitby, “Sometimes it’s hard to be a robot: A call for action on the ethics of abusing artificial agents,” Interacting with Computers, this issue, 2008.

Tools to design, analyse and evaluate user interfaces can be used in user interface design research and in interface modelling research. This demonstration shows two working systems: one in Mathematica that is mathematically sophisticated, and one as a ‘conventional ’ rapid application development environment, where the mathematics is hidden, and which could form the basis of a professional design tool — but which is based rigorously on the same algebraic formalism.

It is usually very hard, both for designers and users, to reason reliably about user interfaces. This article shows that 'push button' and 'point and click' user interfaces are algebraic structures. Users effectively do algebra when they interact, and therefore we can be precise about some important design issues and issues of usability. Matrix algebra, in particular, is useful for explicit calculation and for proof of various user interface properties.With matrix algebra, we are able to undertake with ease unusally thorough reviews of real user interfaces: this article examines a mobile phone, a handheld calculator and a digital multimeter as case studies, and draws general conclusions about the approach and its relevance to design.

Small handheld devices—mobile phones, Pocket PCs etc.—are increasingly being used to access the web. Search engines are the most used web services and are an important factor of user support. Search engine providers have begun to offer their services on the small screen. This paper presents a detailed evaluation of the how easy to use such services are in these new contexts. An experiment was carried out to compare users' abilities to complete search tasks using a mobile phone-sized, handheld computer-sized and conventional, desktop interface to the full Google™ index. With all three interfaces, when users succeed in completing a task, they do so quickly (within 2–3 min) and using few interactions with the search engine. When they fail, though, they fail badly. The paper examines the causes of failures in small screen searching and proposes guidelines for improving these interfaces. In addition, we present and discuss novel interaction schemes that put these guidelines into practice.

Abstract The computer science literature discusses code and algorithms extensively, but not always reliably. Tool support can help ensure integrity between code and explanation so that published papers are more reliable. A versatile, lightweight tool to support explaining code for publication is justified, described and compared with alternatives. The tool works with Java, C and similar languages, and provides support for publishing explanations of real code in LATEX, XML, HTML, etc. Copyright © 2003 John Wiley & Sons, Ltd.

Abstract The Chinese Postman Problem has many applications, including robot exploration, and analysing interactive system and Web site usability. This paper reviews the wide range of applications of the problem and presents complete, executable code to solve it for the case of directed multigraphs. A variation called the ‘open Chinese Postman Problem’ is also introduced and solved. Although optimizations are possible, no substantially better algorithms are likely. Copyright © 2003 John Wiley & Sons, Ltd.

Symmetry is routinely used in visual design, but in fact is not just a visual concept. This paper explores how deeper symmetries in user interface implementations can be 'reflected' in the design of the user interface, and make them easier to use. This deeper application of symmetry for user interface design is related to affordance, and therefore makes that concept constructively applicable. Recommendations for programming better user interfaces are suggested.

Mobile internet technologies, such as WAP, are important for pervasive, anytime, anywhere computing. Although much progress has been made in terms of technological innovation, many of mobile internet systems are difficult to use, lack flexibility and robustness. They give a poor user experience. Evaluation and theoretical analysis of usability combined with innovative design can achieve significant improvements in user performance and satisfaction. Using such multidisciplinary methods explains the negative reactions to WAP, and --- more constructively --- suggest ways of developing more effective and efficient devices and services. Key words: Mobile internet, HCI, Usability, WAP. 1.

Computer viruses are a worrying real-world problem, and a challenge to theoretical modelling. In this issue of the Computer Journal, Erkki Mäkinen proposes universal Turing machines in a critique of an earlier paper, ‘A framework for modelling Trojans and computer virus infection’ (Thimbleby, H., Anderson, S. O. and Cairns, P. (1998) Comp. J., 41, 444–458). This short paper is a reply by those authors.

How hard to users to find interactive devices to use to achieve their goals, and how can we get this information early enough to influence design? We show that Markov modeling can obtain suitable measures, and we provide formulas that can be used for a large class of systems. We analyze and consider alternative designs for various real examples. We introduce a “knowledege/usability graph,” which shows the impact of even a smaller amount of knowledge for the user, and the extent to which designers' knowledge may bias their views of usability. Markov models can be built into design tools, and can therefore be made very convenient for designers to utilize. One would hope that in the future, design tools would include such mathematical analysis, and no new design skills would be required to evaluate devices. A particular concern of this paper is to make the approach accessible. Complete program code and all the underlying mathematics are provided in appendices to enable others to replicate and test all results shown.

Our experience of using Java is disappointing: as a programming language (irrespective of its implementations and libraries), Java itself leaves much to be desired. This paper discusses a few serious problems with Java's design, which leads us to suggest that the language definition should have been an integral part of the design process rather than, as appears, a retrospective commentary. Copyright © 1999 John Wiley & Sons, Ltd.

It is not possible to view a computer operating in the real world, including the possibility of Trojan horse programs and computer viruses, as simply a finite realisation of a Turing machine. We consider the actions of Trojan horses and viruses in real computer systems and suggest a minimal framework for an adequate formal understanding of the phenomena. Some conventional approaches, including biological metaphors, are shown to be inadequate; some suggestions are made towards constructing virally-resistant systems.

Manuals and interactive help are tedious to provide, difficult to maintain, and difficult to ensure that they remain correct, even for simple systems. The result is a loss in product quality, felt particularly by users and designers committed to long-term product development. The paper shows that it is possible to systematically put a system specification and its documentation into exact correspondence. It follows that much previously manual work can be done automatically — and with considerable advantages, including guaranteed correctness and completeness, as well as supporting powerful new features such as intelligent adaptive assistance. This paper shows how interactive assistance can be provided to answer ‘how to?’, ‘why not?’ and other questions.

We discuss and show correct an algorithm that constructs autostereograms. The concern is to develop a correct algorithm that may be implemented in a language such as C or Pascal.

Conventional calculators are badly designed: they suffer from bad computer science — they are unnecessarily difficult to use and bug ridden. I describe a solution, avoiding the problems caused by conventional calculators that is more powerful and arguably much easier to use. The solution has been implemented, and design issues are discussed. This paper shows an interactive system that is declarative, with the advantages of clarity and power that entails. It frees people from working out how a calculation should be expressed to concentrating on what they want solved. An important contribution is to demonstrate the very serious problems users face when using conventional calculators, and hence what a freedom a declarative design brings.

We have built some simple, but useful, cooperative artificial life agents. Based on this experience and by contrasting our work with computer viruses, we argue that artificial life (the simulation of life including evolution) can only remain reliably and indefinitely cooperative if it adheres to explicitly-specified social conventions. Breaking or neglecting these conventions results in systems that are worse than useless; in fact, malicious with respect to human social values.>

The complementary paradigms of text compression and image compression suggest that there may be potential for applying methods developed for one domain to the other. In image coding, lossy techniques yield compression factors that are vastly superior to those of the best lossless schemes and we show that this is also the case for text. This paper investigates the resulting trade-off between subjective quality of the transmission and its compression factor. Two different methods are described, which can be combined into an extremely effective technique that provides far better compression than the present state of the art and yet preserves a reasonable degree of perceived match between the original and received text. The major challenge for lossy text compression is the quantitative evaluation of the quality of this match.

Computerisation is the technological imperative, yet it often results in deeper social malaise than the simpler methods replaced; the alienations, even inefficiencies, that they represent become more rigidly entrenched (Weizenbaum, 1979). All this is well known and widely debated. It is systemic and seems intractable (Illich, 1973). This article discusses a similar problem at the level of user interface design, a

A major factor of system usability is whether the system works at all. This paper discusses bugs and the social environment that allows and encourages them to exist. Many bugs are known about and accepted when software is released to users. They could have been corrected if there had been any motivation to do so. Although individual programmers are often responsible for bugs, various forces within the computing industry, including mistrust of users, drive software manufacturers to strategies that exacerbate the problem. Such methods as software manufacturers adopt ‘in defence’ not only work against users but also undermine scientific work, which in turn retards the advancement of HCI generally.

CSCW (computer-supported cooperative work) is an active research area with many promising applications and benefits. We argue that the plight of the individual user can also be viewed as a CSCW problem, for the individual frequently acts as multiple persona: performing many independent tasks, perhaps in several places. We propose reflexive CSCW to address such issues. Solutions in the reflexive case will of course be of benefit to users even if they are working in a conventional multi-user CSCW context; proposed solutions in CSCW can be re-presented for individual users.

Abstract The inner loop of insert sort can be simplified by using a sentinel value. We suggest a way of avoiding the extra storage normally required for the sentinel, and a way of avoiding the prior, and possibly restrictive, choice of its value. Algorithms are given in Pascal.

The following paper, 'Optimising Self Replicating Programs', would appear on the surface to be a contribution not to be taken too seriously.This is misleading.There is a serious point concerning the need always to analyse fully the logic behind what we are doing.I wish, however, to correct the author on one aspect of the paper; in his references he mentions a page in The Computer Journal, vol.29, no. 6.This particular page is not in the main body of the Journal but is in the Annual Index.The author should note that it is Editorial Board policy not to include in the main body of the Journal pages of the kind to which both he and I refer.Whether or not the reference to vol. 20, no. 4 implies the existence of'virtual journals' and/or 'virtual papers' I must leave to readers to decide: it is presumably a page fault.

Abstract Chareacter‐mapped display terminals very considerably. TIP is a Terminal Independent Package which conceals the details of efficiently handling terminal protocols from applications programs. There is a conflict between the ability to handle esoteric terminal features and providing a uniform programmer's interface for all terminals. This paper discusses the conflicts and the design approch which resolved them.

Cweb is a literate programming system for the programming language C. Experience developing and using it forms the basis of this paper, the purpose of which is to: support Knuth's enthusiasm for literate programming; discuss developments in literate programming support, both within the current framework of cweb and to interactive graphics support; discuss the implementation issues, considering cweb as part of a simple IPSE (Integrated Project Support Environment); the conclusions for IPSEs appear pessimistic.

Many text display editors include ‘word’ function keys. The user effectiveness of such functions depends crucially on a well thought out idea of what ‘words’ are. This paper presents a carefully thought out scheme for word functions and gives the relevant algorithms for them in Pascal. The approach used here has been used with success for well over two years after examining and rejecting many alternatives. It is beyond the scope of this paper to attempt a psychological assertion on the nature of words or word formation.

Abstract Two primitives for structured programming are introduced. The primitives allow a generalized procedure entry and return similar to the ‘loop’ and ‘break’ statements found in many algorithmic languages for control in repetitive commands. Examples are given and the practicality of the primitives especially for interactive programming is stressed. Finally, the detailed implementation of the primitives is discussed; they may be implemented as procedures within an existing language.