Recently, tools for the analysis and visualization of code coverage have become widely available. At first glance, their value in assessing and improving the quality of automated test suites seems to be obvious. Yet, experimental studies as well as experience from projects in industry indicate that their use is not without pitfalls. We found these tools in a number of recent projects quite beneficial. Therefore, we set out to gather code coverage information from one of these projects. In this experience report, first the system under scrutiny as well as our methodology is described. Then, four major questions concerning the impact and benefits of using these tools are discussed. Furthermore, a list of ten lessons learned is derived. The list may help developers judiciously use code coverage tools, in order to reap a maximum of benefits.

This report addresses some of our observations made in a dozen of projects in the area of software testing, and more specifically, in automated testing. It documents, analyzes and consolidates what we consider to be of interest to the community. The major findings can be summarized in a number of lessons learned, covering test strategy, testability, daily integration, and best practices.The report starts with a brief description of five sample projects. Then, we discuss our observations and experiences and illustrate them with the sample projects. The report concludes with a synopsis of these experiences and with suggestions for future test automation endeavors.

This report comprises some experiences, which were made during the development of a point of sale (POS) system. Specific about the project is the fact that it started as an attempt to develop customizable standard software, and then was restructured to deliver a unique project solution. This report details the situation before and after the restructuring, and discusses the experiences made through the ongoing development. These experiences relate mostly to three areas, which are: software project management, prototyping, and testing.

Most current object modeling methods and tools have weaknesses both in the concepts of hierarchical decomposition and in the visualization of these hierarchies. Some methods do not support hierarchical decomposition at all. Those methods which do employ tools that provide explosive zoom as the only means for the visualization of hierarchies. The authors present an approach for the visualization of hierarchical object models based on the notion of fisheye views. This concept integrates local detail and global context of a view in the same diagram and eases navigation in hierarchical structures without offending the principle of abstraction. The work is part of an effort to create a method and language called ADORA that provides strong support for hierarchical decomposition.