arXiv subjects

David Moreno-Lumbreras

Publications and source records attributed to David Moreno-Lumbreras.

5 recordsLinked to original sources

arXiv2026

Not Only for Developers: Exploring Plugin Maintenance for Knowledge-Centric Communities

The adoption of third-party libraries has become integral to modern software development, leading to large ecosystems such as PyPI, NPM, and Maven, where contributors typically share the technical expertise to sustain extensions. In communities that are not exclusively composed of developers, however, maintaining plugin ecosystems can present different challenges. In this early results paper, we study Obsidian, a knowledge--centric platform whose community is focused on writing, organization, and creativity--has built a substantial plugin ecosystem despite not being developer--centric. We investigate what kinds of plugins exist within this hybrid ecosystem and establish a foundation for understanding how they are maintained. Using repository mining and LLM-based topic modeling on a representative sample of 396 plugins, we identify six topics related to knowledge management and tooling, which is (i) dynamic editing and organization, (ii) interface and layouts, (iii) creative writing and productivity, (iv) knowledge sync solutions, (v) linking and script tools, and (vi) workflow enhancements tools. Furthermore, analysis of the Pull Requests from these plugins show that much software evolution has been performed on these ecosystem. These findings suggest that even in mixed communities, plugin ecosystems can develop recognizable engineering structures, motivating future work that highlight three different research directions with six research questions related to the health and sustainability of these non-developer ecosystems.

cs.SE↗

arXiv2026

Understanding Specification-Driven Code Generation with LLMs: An Empirical Study Design

Large Language Models (LLMs) are increasingly integrated into software development workflows, yet their behavior in structured, specification-driven processes remains poorly understood. This paper presents an empirical study design using CURRANTE, a Visual Studio Code extension that enables a human-in-the-loop workflow for LLM-assisted code generation. The tool guides developers through three sequential stages--Specification, Tests, and Function--allowing them to define requirements, generate and refine test suites, and produce functions that satisfy those tests. Participants will solve medium-difficulty problems from the LiveCodeBench dataset, while the tool records fine-grained interaction logs, effectiveness metrics (e.g., pass rate, all-pass completion), efficiency indicators (e.g., time-to-pass), and iteration behaviors. The study aims to analyze how human intervention in specification and test refinement influences the quality and dynamics of LLM-generated code. The results will provide empirical insights into the design of next-generation development environments that align human reasoning with model-driven code generation.

cs.SE↗

arXiv2025

Fuzzy to Clear: Elucidating the Threat Hunter Cognitive Process and Cognitive Support Needs

With security threats increasing in frequency and severity, it is critical that we consider the important role of threat hunters. These highly-trained security professionals learn to see, identify, and intercept security threats. Many recent works and existing tools in cybersecurity are focused on automating the threat hunting process, often overlooking the critical human element. Our study shifts this paradigm by emphasizing a human-centered approach to understanding the lived experiences of threat hunters. By observing threat hunters during hunting sessions and analyzing the rich insights they provide, we seek to advance the understanding of their cognitive processes and the tool support they need. Through an in-depth observational study of threat hunters, we introduce a model of how they build and refine their mental models during threat hunting sessions. We also present 23 themes that provide a foundation to better understand threat hunter needs and suggest five actionable design propositions to enhance the tools that support them. Through these contributions, our work enriches the theoretical understanding of threat hunting and provides practical insights for designing more effective, human-centered cybersecurity tools.

cs.CR↗

arXiv2025

HTML Structure Exploration in 3D Software Cities

Software visualization, which uses data from dynamic program analysis, can help to explore and understand the behavior of software systems. It is common that large software systems offer a web interface for user interaction. Usually, available web interfaces are not regarded in software visualization tools. This paper introduces additions to the web-based live tracing software visualization tool ExplorViz: We add an embedded web view for instrumented applications in the 3D visualization to ease interaction with the given applications and enable the exploration of the thereby displayed HTML content. Namely, the Document Object Model (DOM) is visualized via a three-dimensional representation of the HTML structure in same-origin contexts. Our visualization approach is evaluated in a preliminary user study. The study results give insights into the potential use cases, benefits, and shortcomings of our implemented approach. Based on our study results, we propose directions for further research to support the visual exploration of web interfaces and explore use cases for the combined visualization of software cities and HTML structure. Video URL: https://youtu.be/wBWKlbvzOOE

cs.SE↗

arXiv2021

To VR or not to VR: Is virtual reality suitable to understand software development metrics?

Background/Context: Currently, the usual interface for visualizing data is based on 2-D screens. Recently, devices capable of visualizing data while immersed in VR scenes are becoming common. However, it has not been studied in detail to which extent these devices are suitable for interacting with data visualizations in the specific case of data about software development. Objective/Aim: In this registered report, we propose to answer the following question: "Is comprehension of software development processes, via the visualization of their metrics, better when presented in VR scenes than in 2D screens?" In particular, we will study if answers obtained after interacting with visualizations presented as VR scenes are more or less correct than those obtained from traditional screens, and if it takes more or less time to produce those answers. Method: We will run an experiment with volunteer subjects from several backgrounds. We will have two setups: an on-screen application, and a VR scene. Both will be designed to be as much equivalent as possible in terms of the information they provide. For the former, we use a commercial-grade set of \kibana-based interactive dashboards that stakeholders currently use to get insights. For the latter, we use a set of visualizations similar to those in the on-screen case, prepared to provide the same set of data using the museum metaphor in a VR room. The field of analysis will be related to modern code review, in particular pull request activity. The subjects will try to answer some questions in both setups (some will work first in VR, some on-screen), which will be presented to them in random order. To draw results, we will compare and statistically analyze both the correctness of their answers, and the time spent until they are produced.

cs.SE↗