Iterative improvement was successfully applied to the development of an extensive family of compilers for a family of programming languages ​​on a range of hardware architectures. A set of 17 versions of the system were developed in one place, generating 17 thousand lines of high-level language source code (6500 of executable code). The system was subsequently developed at two different sites, resulting in two different versions of the base language: one version essentially focused on mathematical applications, adding real numbers and various mathematical functions, and the other focused on adding compiler writing capabilities. Each iteration was analyzed from the users' point of view (the language's capabilities were determined in part by the user's needs) and the developer's point of view (the compiler design evolved to be more easily modifiable, for example, to add new data types). Measurements such as coupling and modularization were followed over multiple versions.

Using analytics and measurements as guides for the improvement process is a major difference between iterative improvements and rapid application development, primarily for two reasons:

These measurements and analysis activities can be added to existing rapid development methods.

In fact, the context of multiple iterations brings advantages in the use of measurements. The measures are sometimes difficult to understand at all, although relative changes in the measures through the evolution of the system can be very informative because they provide a basis for comparison. For example, a vector of measurements m1, m2,..., mn can be defined to characterize various aspects of the product at a certain point, such as the total effort made, changes, defects, logical, physical and dynamic attributes, environmental considerations, etc. Thus the observer can tell how product characteristics such as size, complexity, coupling and cohesion increase or decrease over time. They can also monitor the relative change of various aspects of a product or provide limits on measurements to point to potential problems and anomalies.

Contenido

La idea principal detrás de mejoramiento iterativo es desarrollar un sistema de manera incremental, permitiéndole al desarrollador sacar ventaja de lo que se ha aprendido a lo largo del desarrollo anterior, incrementando, versiones entregables del sistema. El aprendizaje viene de dos vertientes: el desarrollo del sistema, y su uso (mientras sea posible). Los pasos claves en el proceso son comenzar con una implementación simple de los requerimientos del sistema, e iterativamente mejorar la secuencia evolutiva de versiones hasta que el sistema completo esté implementado. En cada iteración, se realizan cambios en el diseño y se agregan nuevas funcionalidades y capacidades al sistema.

Básicamente este modelo se basa en dos premisas:.

El proceso en sí mismo consiste de:.

Considerations on the timing of application

To integrate usability into a development process, it is not enough to assign usability techniques to development activities, since not all usability techniques are applicable at any time in an iterative development.

For example, techniques for developing the product concept are intended for application in early development efforts, when needs are identified and the general outline of the system is established. Although it is advisable to also apply them later, to refine the concept, their main application effort is in the initial development tasks.[5]​.

Initialization stage

A version of the system is created. The goal of this stage is to create a product with which the user can interact, and therefore provide feedback to the process. It should offer a snapshot of the key aspects of the problem and provide a solution simple enough to be easily understood and implemented. To guide the iteration process, a project checklist is created, which contains a history of all the tasks that need to be performed. It includes things like new functionality to be implemented, and areas of redesign of the existing solution. This checklist is periodically and constantly reviewed as a result of the analysis phase..

Iteration stage

This stage involves redesigning and implementing a project checklist task, and analyzing the most recent version of the system. The goal of the design and implementation of any iteration is to be simple, direct and modular, in order to support the redesign of the stage or as a task added to the project checklist. The code can, in certain cases, represent the largest source of system documentation. The analysis of an iteration is based on user feedback and the analysis of the available functionalities of the program. It involves the analysis of structure, modularity, usability, reliability, efficiency and effectiveness (achieving goals). The project checklist is modified in light of the analysis results.

Primary guidelines guiding implementation and analysis include:.

Iterative improvement was successfully applied to the development of an extensive family of compilers for a family of programming languages ​​on a range of hardware architectures. A set of 17 versions of the system were developed in one place, generating 17 thousand lines of high-level language source code (6500 of executable code). The system was subsequently developed at two different sites, resulting in two different versions of the base language: one version essentially focused on mathematical applications, adding real numbers and various mathematical functions, and the other focused on adding compiler writing capabilities. Each iteration was analyzed from the users' point of view (the language's capabilities were determined in part by the user's needs) and the developer's point of view (the compiler design evolved to be more easily modifiable, for example, to add new data types). Measurements such as coupling and modularization were followed over multiple versions.

Using analytics and measurements as guides for the improvement process is a major difference between iterative improvements and rapid application development, primarily for two reasons:

These measurements and analysis activities can be added to existing rapid development methods.

In fact, the context of multiple iterations brings advantages in the use of measurements. The measures are sometimes difficult to understand at all, although relative changes in the measures through the evolution of the system can be very informative because they provide a basis for comparison. For example, a vector of measurements m1, m2,..., mn can be defined to characterize various aspects of the product at a certain point, such as the total effort made, changes, defects, logical, physical and dynamic attributes, environmental considerations, etc. Thus the observer can tell how product characteristics such as size, complexity, coupling and cohesion increase or decrease over time. They can also monitor the relative change of various aspects of a product or provide limits on measurements to point to potential problems and anomalies.