Journal of Systems and Software最新文献

Grover’s algorithm is a well-known contribution to quantum computing. It searches one value within an unordered sequence faster than any classical algorithm. A fundamental part of this algorithm is the so-called oracle, a quantum circuit that marks the quantum state corresponding to the desired value. A generalisation of it is the oracle for Amplitude Amplification, that marks multiple desired states. In this work we present a classical algorithm that builds a phase-marking oracle for Amplitude Amplification. This oracle performs a less-than operation, marking states representing natural numbers smaller than a given one. Results of both simulations and experiments are shown to prove its functionality. This less-than oracle implementation works on any number of qubits and does not require any ancilla qubits. Regarding depth, the proposed implementation is compared with the one generated by Qiskit automatic method, Diagonal. We show that the depth of our less-than oracle implementation is always lower. In addition, a comparison with another method for oracle generation in terms of gate count is also conducted to prove the efficiency of our method. The result presented here is part of a research work that aims to achieve reusable quantum operations that can be composed to perform more complex ones. The final aim is to provide Quantum Developers with tools that can be easily integrated in their programs/circuits.

With the recent advancement of Artificial Intelligence (AI) and Large Language Models (LLMs), AI-based code generation tools become a practical solution for software development. GitHub Copilot, the AI pair programmer, utilizes machine learning models trained on a large corpus of code snippets to generate code suggestions using natural language processing. Despite its popularity in software development, there is limited empirical evidence on the actual experiences of practitioners who work with Copilot. To this end, we conducted an empirical study to understand the problems that practitioners face when using Copilot, as well as their underlying causes and potential solutions. We collected data from 473 GitHub issues, 706 GitHub discussions, and 142 Stack Overflow posts. Our results reveal that (1) Operation Issue and Compatibility Issue are the most common problems faced by Copilot users, (2) Copilot Internal Error, Network Connection Error, and Editor/IDE Compatibility Issue are identified as the most frequent causes, and (3) Bug Fixed by Copilot, Modify Configuration/Setting, and Use Suitable Version are the predominant solutions. Based on the results, we discuss the potential areas of Copilot for enhancement, and provide the implications for the Copilot users, the Copilot team, and researchers.

With the deployment of Deep Neural Network (DNN) systems in security-critical fields, more and more researchers are concerned about DNN robustness. Unfortunately, DNNs are vulnerable to adversarial attacks and produce completely wrong outputs. This inspired numerous testing works devoted to improving the adversarial robustness of DNNs. Coverage and uncertainty criteria were proposed to guide sample selections for DNN retraining. However, they are greatly limited to evaluating DNN abnormal behaviors rather than locating the root cause of adversarial vulnerability. This work aims to bridge this gap. We propose an adversarial testing framework, DeepFeature, using robust features. DeepFeature generates robust features related to the model decision-making. It locates the weak features within these features that fail to be transformed by the DNN. They are the main culprits of vulnerability. DeepFeature selects diverse samples containing weak features for adversarial retraining. Our evaluation shows that DeepFeature significantly improves overall robustness, average improved by 77.83%, and individual robustness, average improved by 42.81‰, of the models in adversarial testing. Compared with coverage and uncertainty criteria, these two performances are improved by 3.93% and 15.00% in DeepFeature, respectively. The positive correlation coefficient between DeepFeature and improved robustness can achieve 0.858, and the $p$-value is 0.001.

Software development requires collaborative efforts and consensus among developers, emphasizing the need for effective communication and knowledge sharing within the teams. In line with this, GitHub introduced GitHub Discussions, a collaborative communication feature for the community around an open-source or internal project. The user-friendly interface of Discussions facilitates easy sharing of links to different resources. Maintainers highlight this feature’s significance, enabling users to maintain a well-informed project environment. We hypothesize that link-sharing activities in Discussions contribute to disseminating project knowledge. To investigate this hypothesis, we conducted a mixed-method study combining qualitative and quantitative analysis based on a convenience sample of ten open-source projects. We aimed to gain insight into the scope and intentions behind these shared links. Our findings indicate that link-sharing activities are common in the Discussions. Users share links to resources directly or indirectly related to the project/repository. Discussions users share links to project documentation, source code, and issues, aiming to clarify concepts, provide supplementary information, offer context to questions, and suggest new features. These findings offer insights for project maintainers to understand Discussions usage better and enable the GitHub Engineering team to promote the feature adoption.

Legacy software poses a critical challenge for organizations due to the costs of maintaining and modernizing outdated systems, as well as the scarcity of experts in aging programming languages. The issue extends beyond commercial applications, affecting public administration, as exemplified by the urgent need for COBOL programmers during the COVID-19 pandemic. In response, this work introduces a modernization approach based on dynamic language product lines, a subset of dynamic software product lines. This approach leverages open language implementations and dynamically generated micro-languages for the incremental migration of legacy systems to modern technologies. The language can be reconfigured at runtime to adapt to the execution of either legacy or modern code, and to generate a compatibility layer between the data types handled by the two languages. Through this process, the costs of modernizing legacy systems can be spread across several iterations, as developers can replace legacy code incrementally, with legacy and modern code coexisting until a complete refactoring is possible. By moving the overhead of making legacy and modern features work together in a hybrid system from the system implementation to the language implementation, the quality of the system itself does not degrade due to the introduction of glue code. To demonstrate the practical applicability of this approach, we present a case study on a COBOL system migration to Java. Using the Neverlang language workbench to create modular and reconfigurable language implementations, both the COBOL interpreter and the application evolve to spread the development effort across several iterations. Through this study, this work presents a viable solution for organizations dealing with the complexity of modernizing legacy software to contemporary technologies. The contributions of this work are (i) a language-oriented, incremental refactoring process for legacy systems, (ii) a concrete application of open language implementations, and (iii) a general template for the implementation of interoperability between languages in hybrid systems.

Context:

The Open Science (OS) movement promotes the value of making public the research artifacts (datasets, analysis scripts, guidelines, etc.) used during empirical studies. OS is widely known in areas such as Medicine or Biology, where the process of sharing research artifacts is subject to strict protocols. Unfortunately, in Software Engineering (SE), this process is carried out in a non-systematic way, resulting in incomplete or inaccurate material shared by researchers, which hinders the reproducibility and replicability of empirical studies. Nevertheless, in recent years, it seems that the Empirical Software Engineering (ESE) community is embracing some of the proposed OS initiatives, such as the one proposed by the Association for Computing Machinery (ACM), which provides a badge system to evaluate the quality of a research artifact. This badge system has been adopted by several SE conferences as a method of assessing research artifacts.

Aims:

Focusing on human-oriented experiments (HOEs) in SE, whose research artifacts are more complex than those for computational experiments, this work applies Design Science Research (DSR) with a twofold purpose: (i) review the current status of HOEs research artifacts publication through evaluation of this practice in the most relevant ESE journals , and (ii) propose a structured outline for HOEs research artifacts driven by the aforementioned ACM badging policy.

Method:

Regarding the first purpose, we carried out a survey to analyze the current status of the publication of research artifacts considering relevant peer review journals and the quality of 106 research artifacts published in these journals with respect to the ACM badging policy. For the second purpose, an iterative process was carried out to review the content of 106 research artifacts research and their concordance with ACM badges, obtaining a structured scheme for HOEs research artifacts that has been validated through a detailed review of 12 research artifacts obtained from some of those of ACM badges in relevant SE conferences. In addition, we validated the proposal in the research artifacts of 2 of our own experiments.

Results:

Our survey reveals issues such as the 39,70% of journal studies making completely accessible their research artifacts; most of the analyzed research artifacts are incomplete; the most common repositories used in the ESE community to share the research artifacts are GitHub, institutional repositories, and Zenodo. On the other hand, the validated and structured research artifact outline consists of a list of ordered sections containing a set of artifacts, which can be mandatory or not to achieve a particular ACM badge. For its internal validation, several improvement iterations on the first release of the outline have been carried out based on the conference guidelines, the ACM badging policy, and other relevant proposals.