Ernst Denert Award for Software Engineering 2022

The Ernst Denert Award is already existing since 1992, which does not only honor the award winners but also the software engineering field in total. Software engineering is a vivid and intensively extending field that regularly spawns new subfields such as automotive software engineering, research software engineering, or quantum software engineering, covering specific needs but also generalizing solutions, methods, and techniques when they become applicable. This is the introductory chapter of the book on the Ernst Denert Software Engineering Award 2022. It provides an overview of the five nominated PhD theses.

This thesis constitutes the first work in the RE community that studies the potential of extracting conditional statements from requirements. It is intended to stimulate further engagement of researchers and practitioners in the field of conditionals in RE artifacts. In essence, we present fundamental research on the notion of conditionals in requirements as well as methods for their fine-grained extraction. We show that conditionals are prevalent in requirements and mainly occur in explicit, marked form. Further, we reveal that conditionals are a source of ambiguity, and there is not just one way to interpret them formally. This affects any automated analysis that builds upon formalized requirements (e.g., inconsistency checking) and may also influence guidelines for writing requirements. We also present our tool-supported approach CiRA, capable of detecting conditionals in NL requirements and extracting them in fine-grained form. We evaluate our approach in a case study with three industry partners, namely, Allianz Deutschland AG (insurance), Ericsson (telecommunication), and Leopold Kostal GmbH & Co. KG (automotive), and highlight that automated conditional extraction facilitates automated acceptance test creation. CiRA is available at http://​www.​cira.​bth.​se/​demo/​.

To ensure the security of a software system, it is vital to keep up with changing security precautions, attacks, and mitigations. Although model-based development enables addressing security already at design-time, design models are often inconsistent with the implementation or among themselves. Such inconsistencies hinder the effective realization and verification of secure software systems. In addition, variants of software systems are another burden to developing secure systems. Vulnerabilities must be identified and fixed on all variants or else attackers could be well-guided in attacking unfixed variants. To ensure security in this context, in the thesis (Peldszus, Security Compliance in Model-driven Development of Software Systems in Presence of Long-Term Evolution and Variants. Springer, Berlin; 2022), we present GRaViTY, an approach that allows security experts to specify security requirements on the most suitable system representation. To preserve security, based on continuous automated change propagation, GRaViTY automatically checks all system representations against these security requirements. To systematically improve the object-oriented design of a software-intensive system, GRaViTY provides security-preserving refactorings. For both continuous security compliance checks and refactorings, we show the application to variant-rich software systems. To support legacy systems, GRaViTY allows to automatically reverse-engineer variability-aware UML models and semi-automatically map existing design models to the implementation. Besides evaluations of the individual contributions, we demonstrate applicability of the approach in two real-world case studies, the iTrust electronics health records system and the Eclipse Secure Storage. This book chapter provides a summary of the thesis, focusing on the addressed problems, identified and answered research questions, the general solution, and its application of it to two case studies. For details on the individual solutions, please refer to the thesis and the corresponding publications referenced in this book chapter.

This chapter presents LEMMA (Language Ecosystem for Modeling Microservice Architecture). LEMMA enables the application of Model-Driven Engineering (MDE) to Microservice Architecture (MSA). LEMMA mitigates the complexity of MSA by decomposing it along four viewpoints on microservice architectures, each capturing the concerns of different MSA stakeholders in dedicated architecture models. LEMMA formalizes the syntax and semantics of these models with specialized modeling languages that are integrated based on an import mechanism, thus enabling holistic MSA modeling. LEMMA also bundles its own model processing framework (MPF) to facilitate model processor implementation for technology-savvy MSA stakeholders without a background in MDE.

We describe the design and development of LEMMA and exemplify the usage of its modeling languages and MPF for a case study microservice architecture. In addition, we present practical applications of LEMMA for microservice code generation, architecture reconstruction, quality analysis, defect resolution, and establishing a common architecture understanding. A comparison of LEMMA with related approaches reveals that LEMMA has particular strengths in (i) language-level extensibility, allowing model-based reification of architectural patterns; (ii) model processing, by bundling sophisticated code generators and static quality analyzers; and (iii) versatility, making LEMMA applicable in microservice development, operation, architecture reconstruction, and quality assessment.

Automated Static Analysis Tools (ASATs) are an additional tool available to developers in their pursuit of high-quality software. ASATs match source code against configured rules and produce a warning when a rule is violated. However, the evaluation of the warnings by developers as well as the resolution of warnings requires time. This raises the question of whether we are able to evaluate the usefulness of ASATs empirically. Within this chapter, we present the results of four case studies, which investigate different aspects regarding the impact of ASATs on software quality and the perception of the developers thereof. We present results regarding the evolution of ASAT warnings from a longitudinal study of 54 open-source projects. To evaluate the impact on defects, we present results from two studies. The first study is evaluating predictive models in the context of defect prediction with ASAT-based features. The second study provides a statistical investigation of the differences between changes that induce a defect and all other changes. In order to observe the developer’s perspective regarding ASAT warnings and other software quality metrics, we include the results of a study of developer intent, which compares changes where the developers intend to improve the quality of the code base with all other changes to see which quality metrics and ASAT warnings change in which way. We employ methods of empirical software engineering research to investigate these relationships and provide evidence-based information for researchers and practitioners alike. Within our studies, we can show empirically that we are able to measure an impact on quality. However, the effect is surprisingly small. Moreover, our investigation of developer intents yield information about the magnitude of bug fixing as a driver for complexity in software. Our results can help practitioners estimate the possible impact of introducing an ASAT on defects, as well as provide guidelines for managing the complexity of software.