Harald Rønneberg

A case study comparing defect profiles of a reused framework and of applications reusing it

The benefits of software reuse have been studied for many years. Several previous studies have observed that reused software has a lower defect density than newly built software. However, few studies have investigated empirically the reasons for this phenomenon. To date, we have only the common sense observation that as software is reused over time, the fixed defects will accumulate and will result in high-quality software. This paper reports on an industrial case study in a large Norwegian Oil and Gas company, involving a reused Java class framework and two applications that use that framework. We analyzed all trouble reports from the use of the framework and the applications according to the Orthogonal Defect Classification (ODC), followed by a qualitative Root Cause Analysis (RCA). The results reveal that the framework has a much lower defect density in total than one application and a slightly higher defect density than the other. In addition, the defect densities of the most severe defects of the reused framework are similar to those of the applications that are reusing it. The results of the ODC and RCA analyses reveal that systematic reuse (i.e. clearly defined and stable requirements, better design, hesitance to change, and solid testing) lead to lower defect densities of the functional-type defects in the reused framework than in applications that are reusing it. However, the different “nature” of the framework and the applications (e.g. interaction with other software, number and complexity of business logic, and functionality of the software) may confound the causal relationship between systematic reuse and the lower defect density of the reused software. Using the results of the study as a basis, we present an improved overall cause–effect model between systematic reuse and lower defect density that will facilitate further studies and implementations of software reuse.

The Impact of Test Driven Development on the Evolution of a Reusable Framework of Components An Industrial Case Study

Test driven development (TDD) is a software engineering technique to promote fast feedback, task-oriented development, improved quality assurance and more comprehensible low-level software design. Benefits have been shown for non-reusable software development in terms of improved quality (e.g. lower defect density). We have carried out an empirical study of a framework of reusable components, to see whether these benefits can be shown for reusable components. The framework is used in building new applications and provides services to these applications during runtime. The three first versions of this framework were developed using traditional test-last development, while for the two latest versions TDD was used. Our results show benefits in terms of reduced mean defect density (35.86%), when using TDD, over two releases. Mean change density was 76.19% lower for TDD than for test-last development. Finally, the change distribution for the TDD approach was 33.3% perfective, 5.6% adaptive and 61.1% preventive.

A text mining approach to integrating business process models and governing documents

As large companies are building up their enterprise architecture solutions, they need to relate business process descriptions to lengthy and formally structured documents of corporate policies and standards. However, these documents are usually not specific to particular tasks or processes, and the user is left to read through a substantial amount of irrelevant text to find the few fragments that are relevant to him. In this paper, we describe a text mining approach to establishing links between business process model elements and relevant parts of governing documents in Statoil, one of Norway’s largest companies. The approach builds on standard IR techniques, gives us a ranked list of text fragments for each business process activity, and can easily be integrated with Statoil’s enterprise architecture solution. With these ranked lists at hand, users can easily find the most relevant sections to read before carrying out their activities.

Using domain-driven design to evaluate commercial off-the-shelf software

Purchasing a Commercial-Off-The-Shelf (COTS) package solution can be a complex and daunting task. Selecting and evaluating the right candidate is difficult, especially when the solution aims at the heart of company business. The companys competitive edge must be maintained, while at the same time ensuring the intended goals such as reduced costs and better functional coverage. A good Enterprise Architecture should be a prime tool when evaluating several solutions against the companys needs.In this paper we will recount the experience and lessons learned when we evaluated three COTS systems to replace a set of legacy oil trading and operations systems. Based on weaknesses in our Enterprise Architecture, we applied strategic domain-driven design principles to extend our Enterprise Architecture during the evaluation. We found that these techniques enabled us to thoroughly analyse our domain with the domain experts and provide answers based on tacit domain knowledge, without going through the cost and effort of performing a full-scale architectural analysis. At the same time, the tacit domain knowledge became explicit and shared, easing the communication with various stakeholders.

Change profiles of a reused class framework vs. two of its applications

Software reuse is expected to improve software productivity and quality. Although many empirical studies have investigated the benefits and challenges of software reuse from development viewpoints, few studies have explored reuse from the perspective of maintenance. This paper reports on a case study that compares software changes during the maintenance and evolution phases of a reused Java class framework with two applications that are reusing the framework. The results reveal that: (1) The reused framework is more stable, in terms of change density, than the two applications that are reusing it. (2) The reused framework has profiles for change types that are similar to those of the applications, where perfective changes dominate. (3) The maintenance and evolution lifecycle of both the reused framework and its applications is the same: initial development, followed by a stage with extending capabilities and functionality to meet user needs, then a stage in which only minor defect repairs are made, and finally, phase-out. However, the reused framework goes faster from the stage of extending capabilities to the stage in which only minor defect repairs are made than its applications. (4) We have validated that several factors, such as are functionalities, development practice, complexity, size, and age, have affected the change densities and change profiles of the framework and applications. Thus, all these factors must be considered to predict change profiles in the maintenance and evolution phase of software.

A Case Study of Defect-Density and Change-Density and their Progress over Time

We have performed an empirical case study, investigating defect-density and change-density of a reusable framework compared with one application reusing it over time at a large Oil and Gas company in Norway, Statoil ASA. The framework, called JEF, consists of seven components grouped together, and the application, called DCF, reuses the framework, without modifications to the framework. We analyzed all trouble reports and change requests from three releases of both. Change requests in our study covered any changes (not correcting defects) in the requirements, while trouble reports covered any reported defects. Additionally, we have investigated the relation between defect-density and change-density both for the reusable JEF framework and the application. The results revealed that the defect-density of the reusable framework was lower than the application. The JEF framework had higher change-density in the first release, but lower change-density than the DCF application over the successive releases. For the DCF application, on the other hand, a slow increase in change-density appeared. On the relation between change-density and defect-density for the JEF framework, we found a decreasing defect-density and change-density. The DCF application here showed a decreasing defect-density, with an increasing change-density. The results show that the quality of the reusable framework improves and it becomes more stable over several releases, which is important for reliability of the framework and assigning resources

Architectural improvement by use of strategic level domain-driven design

In this paper we present the experience gained and lessons learned when the IT department at Statoil ASA, a large Oil and Gas company in Norway, extended their Enterprise Architecture with strategic level Domain-Driven design techniques and used the extended Enterprise Architecture to improve the software architecture of a large enterprise system.Traditionally, Enterprise Architecture has been prescribed as the key tool to conquer complexity and align IT development with business priorities and strategies, but we found our Enterprise Architecture too coarse to be practical useful at the software level.By extending our Enterprise Architecture with context maps and the process of context mapping valuable insight was gained, insight that enabled better scoping of new projects and architectural improvement of existing software in a controlled way.In addition, use of responsibility layers combined with context maps reduces the perceived complexity of the architecture. Use of other techniques such as distillation and identification of the core domain looks promising at the tactical level of a single project, but its value is more uncertain at the strategic level.The key issue is that large enterprise systems do not have a single core. On the other hand, at the project level, there should always be a core, and the project is best of by knowing its core domain and aim its best resources to work with the core.

Experience Report on the Effect of Software Development Characteristics on Change Distribution

This paper reports on an industrial case study in a large Norwegian Oil and Gas company (StatoilHydro ASA) involving a reusable Java-class framework and an application that uses that framework. We analyzed software changes from three releases of the framework and the application. On the basis of our analysis of the data, we found that perfective and corrective changes account for the majority of changes in both the reusable framework and the non-reusable application. Although adaptive changes are more frequent and has longer active time in the reusable framework, it went through less refactoring compared to the non-reusable application. For the non-reusable application we saw preventive changes as more frequent and with longer active time. We also found that designing for reuse seems to lead to fewer changes, as well as we saw a positive effect on doing refactoring.