Andrew Forward

How software engineers use documentation: the state of the practice

Software engineering is a human task, and as such we must study what software engineers do and think. Understanding the normative practice of software engineering is the first step toward developing realistic solutions to better facilitate the engineering process. We conducted three studies using several data-gathering approaches to elucidate the patterns by which software engineers (SEs) use and update documentation. Our objective is to more accurately comprehend and model documentation use, usefulness, and maintenance, thus enabling better decision making and tool design by developers and project managers. Our results confirm the widely held belief that SEs typically do not update documentation as timely or completely as software process personnel and managers advocate. However, the results also reveal that out-of-date software documentation remains useful in many circumstances.

The relevance of software documentation, tools and technologies: a survey

This paper highlights the results of a survey of software professionals. One of the goals of this survey was to uncover the perceived relevance (or lack thereof) of software documentation, and the tools and technologies used to maintain, verify and validate such documents. The survey results highlight the preferences for and aversions against software documentation tools. Participants agree that documentation tools should seek to better extract knowledge from core resources. These resources include the systems source code, test code and changes to both. Resulting technologies could then help reduce the effort required for documentation maintenance, something that is shown to rarely occur. Our data reports compelling evidence that software professionals value technologies that improve automation of the documentation process, as well as facilitating its maintenance.

Problems and opportunities for model-centric versus code-centric software development: a survey of software professionals

We present some results of a survey of 113 software practitioners conducted between April and December 2007. The aim of the survey was to uncover their attitudes and experiences regarding software modeling, and development approaches that avoid modeling. We were motivated by observations that modeling is not widely adopted; many developers continue to take a code-centric approach. Key findings overall include: Modeling tools are primarily used to create documentation and for up-front design with little code generation; and participants believe that model-centric approaches to software engineering are easier but are currently not very popular as most participants currently work in code-centric environments. Key findings from sub-samples include: problems identified with model-centric approaches are similar regardless of a participants country. Programmers that model extensively (versus those that do not model much) are more likely to agree that models become out of date and inconsistent with code.

A taxonomy of software types to facilitate search and evidence-based software engineering

Empirical software research could be improved if there was a systematic way to identify the types of software for which empirical evidence applies. This is because results are unlikely to be globally applicable, but are more likely to apply only in certain contexts such as the type of software on which the evidence has been tested. We present a software taxonomy that should help researchers to apply their research systematically to particular types of software. The taxonomy was generated using existing partial taxonomies and input from survey participants. If a taxonomy such as ours gains acceptance, it will facilitate comparison and appropriate application of research. In the paper, we present the benefits of such a taxonomy, the process we used to develop it, and the taxonomy itself.

Model-driven rapid prototyping with Umple

The emergence of model‐driven software development brings new opportunities and challenges for rapid prototyping. On the one hand, the modeling process is inherently abstract, removing the prototyper from details, and letting him or her focus on exploring design alternatives for various aspects of the system. On the other hand, the most popular modeling languages and tools entirely omit the modeling and generating of user interfaces. As a result, the benefit of user interface prototypes as a medium for interaction with the user and customer is lost. This paper presents a model‐oriented technology called Umple that can be used for prototyping and also supporting model driven engineering. Umple allows end users to quickly create class and state machine models and to incrementally embed implementation artifacts. At any point in the modeling process, users can quickly generate a fully functional prototype that exposes modeling implications on the user interface, and allows stakeholders to get a feel of how the full system will behave. Copyright

Teaching UML using umple: Applying model-oriented programming in the classroom

We show how a technology called Umple can be used to improve teaching UML and modeling. Umple allows UML to be viewed both textually and graphically, with updates to one view reflected in the other. It allows UML concepts to be added to programming languages, plus web-based code generation from UML to those languages. We have used Umple in student laboratories and assignments for two years, and also live in the classroom. In a survey, students showed enthusiasm about Umple, and indicated they believe it helps them understand UML better. Improvements in their grades also support our approach.

Improving program comprehension by enhancing program constructs: An analysis of the Umple language

Umple is a set of extensions to object-oriented languages that provides a concrete syntax for UML abstractions like associations. We argue that Umple will help increase program comprehension by allowing developers to describe a system at a more abstract level, and by reducing the volume of code.

Umplification: Refactoring to Incrementally Add Abstraction to a Program

Umple adds UML abstractions to base programming languages such as Java, PHP and Ruby. The resulting program can be seen as both model and code at the same time. Umplification is the process of converting a base language program to Umple by a series of refactorings. The result is a program semantically equivalent to the original, but which can be rendered and edited as UML diagrams by a modeling tool or manipulated as a textual program for those who prefer. In this paper we discuss the principles of Umple, the process of umplification, and two examples including umplification of the Umple compiler itself.

Improving Code Generation for Associations: Enforcing Multiplicity Constraints and Ensuring Referential Integrity

UML classes involve three key elements: attributes, associations, and methods. Current object-oriented languages, like Java, do not provide a distinction between attributes and associations. Tools that generate code from associations currently provide little support for the rich semantics available to modellers such as enforcing multiplicity constraints or maintaining referential integrity. In this paper, we introduce a syntax for describing associations using a model-oriented language called Umple. We show source code from existing code-generation tools and highlight how the issues above are not adequately addressed. We outline code generation patterns currently available in Umple that resolve these difficulties and address the issues of multiplicity constraints and referential integrity.

Exploring a Model-Oriented and Executable Syntax for UML Attributes

Implementing UML attributes directly in an object-oriented language may not appear to be complex, since such languages already support member variables. The distinction arises when considering the differences between modelling a class and implementing it. In addition to representing attributes, member variables can also represent association ends and internal data including counters, caching, or sharing of local data. Attributes in models also support additional characteristics such as being unique, immutable, or subject to lazy instantiation. In this paper we present modeling characteristics of attributes from first principles and investigate how attributes are handled in several open-source systems. We look code-generation of attributes by various UML tools. Finally, we present our own Umple language along with its code generation patterns for attributes, using Java as the target language.