第9章ソフトウェア工学

Software has become an essential part of our lives. It has revolutionized how we communicate, work, learn, and entertain ourselves. From the smartphones in our pockets to the complex systems that power hospitals and transportation networks, software is everywhere.

Because the focus of software engineering is to build software, it is important to understand what software is. A piece of software is made up of instructions that can be executed and generally includes documentation that describes the software operation and use. These instructions tell computers what to do. They function like a recipe that tells a chef how to cook a meal. Software is designed to accomplish a specific objective or purpose. Word processing software has the objective or purpose to help people create documents. A game has the objective or purpose of providing entertainment. The software built into an automobile has many purposes, including helping a driver drive the car, helping the owner maintain the car, and helping a driver and/or passengers avoid getting lost.

Software is generally created based on a set of requirements, and although software does not wear out, these requirements often change over time. Software must, therefore, be updated often to respond to these changes. In addition, software can contain defects (i.e., bugs), which are expected to be fixed in a reasonable time after they are reported.

Over the years, software engineering researchers and practitioners have converged on defining a generic software engineering process model, or process framework, that can be used as a template to characterize the generic activities performed by all the software engineering process models that are used to support the software development life cycle (SDLC). The SDLC is a structured set of the framework activities required to develop a software solution based on a set of requirements. A process framework generally encompasses four framework (or generic) activities: inception, elaboration, construction, and deployment. These activities are also known as phases.

You’ll learn more about each of these phases later in the chapter. In the meantime, it should be noted that the deployment phase covers integration and user acceptance testing, installation, and maintenance/support activities. This means that maintenance is an integral part of the software engineering process. In fact, the cost of software maintenance/continuous deployment today often exceeds the cost of software construction, especially if software remains in use for a long time.

There are three main categories of software. Software that enables you to control hardware and provides an environment in which other software can run is called system software. An example of this is an operating system, such as Microsoft Windows, Android, and MacOS, which runs on a laptop or a mobile phone and typically controls and provides access to the computer’s basic functionality. This system software also includes driver software, which is the code used by hardware devices to interact with the operating system.

Software that enables you to fulfill common tasks, such as creating a text document, drawing a picture, or playing music, is called application software. Examples include Microsoft Word and Paint, MacOS Pages, and iOS Files. Application software may include specialized categories of software such as scientific (or engineering) software, which is software that aids you in solving mathematical, scientific, and engineering problems, such as finding the roots of an equation. It can also include software services, which are software programs that provide specific responses or actions based on a request, such as returning stock quotes, obtaining weather conditions, or finding what current jobs are available.

Software that is integrated with hardware and can include both application and system software features is called embedded software. Examples include software in a smartwatch, an automobile control system, or microwave. Interaction with embedded software is usually limited, and users are often unaware that they are interacting with software. With the increased popularity of the Internet of Things (IoT), embedded software development is increasing.

Software engineers must often deal with legacy software that has been written in the past, relies on obsolete technology, and is still in use today. This software presents special challenges to those who maintain it, especially when they need to adapt it to a new computing environment or technology and when they are to extend its functionality so that it is interoperable with new software systems. The need for software maintenance of legacy software stems from the rapid changes that have taken place in hardware. For example, a common smartphone contains a more powerful processor and more memory than computers that were used just twenty years ago. Changes in hardware are accompanied by changes in software technologies, and these are quickly adopted by users who are then reluctant to use legacy software in its original shape. For example, the graphical user interface of a software system that was developed twenty years ago now looks outdated and leads to an unpleasant user experience.

There have also been dramatic changes in processor technology. Twenty years ago, most processors consisted of a single core, while now it is common for a processor to have multiple cores and coprocessors. This requires changes in how software is programmed because software that can utilize multiple processor cores has a competitive advantage compared with software that uses only a single core. To support these new capabilities, major programming languages have evolved toward supporting multithreading, which makes maintenance of a legacy code base even more challenging.