Development and Adoption of SATD Detection Tools: A State-of-practice Report
Edi Sutoyo, Andrea Capiluppi · 2024
Read Development and Adoption of SATD Detection Tools: A State-of-practice Report by Edi Sutoyo, Andrea Capiluppi (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-14217.txt.
Introduction
Developers often leave notes in code admitting a shortcut or a flaw, a practice called self-admitted technical debt. Software tools exist to detect these admissions, but how healthy is that tooling itself? Edi Sutoyo and Andrea Capiluppi survey the literature and the available tools to report on the state of practice. They check how accessible the tools are and whether they still work as platforms change around them.
Many tools, they find, suffer from outdated technology, poor upkeep, and narrow platform support. Only a few are actively maintained, which blocks wider adoption in real projects. The report names common anti-patterns in how such tools get built and suggests corrections. The authors urge builders to follow FAIR principles, making tools findable, accessible, interoperable, and reusable. They also call for closer teamwork between universities and industry to keep detection tools alive and useful.
Catalog record
| CREATOR | Edi Sutoyo, Andrea Capiluppi |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-14217
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Development and Adoption of SATD Detection Tools: A State-of-practice Report Edi Sutoyo, Andrea Capiluppi Abstract Self-Admitted Technical Debt (SATD) refers to instances where developers knowingly introduce suboptimal solutions into code and document them, often through textual artifacts. This paper provides a comprehensive state-of-practice report on the development and adoption of SATD detection tools. Through a systematic review of the available literature and tools, we examined their overall accessibility. Our findings reveal that, although SATD detection tools are crucial for maintaining software quality, many face challenges such as technological obsolescence, poor maintenance, and limited platform compatibility. Only a small number of tools are actively maintained, hindering their widespread adoption. This report discusses common anti-patterns in tool development, proposes corrections, and highlights the need for implementing Findable, Accessible, Interoperable, and Reusable (FAIR) principles and fostering greater collaboration between academia and industry to ensure the sustainability and efficacy of these tools. The insights presented here aim to drive more robust management of technical debt and enhance the reliability of SATD tools. [orcid=0000-0002-8413-5070, [email protected], ] [orcid=0000-0001-9469-6050, [email protected], ] keywords self-admitted technical debt,SATD,detection tools,FAIR principles,software engineering practices † † copyrightyear: 2024 † † copyright: Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). † † venue: BENEVOL24: The 23rd Belgium-Netherlands Software Evolution Workshop, November 21-22, Namur, Belgium † † address: Bernoulli Institute, University of Groningen, Groningen, The Netherlands † † address: Department of Information Systems, Telkom University, Bandung, Indonesia 1 Introduction Technical Debt (TD) broadly refers to suboptimal code or design choices that compromise long-term software maintainability Cunningham 1992 . Within this domain, self-admitted technical debt (SATD) refers to instances where developers knowingly document suboptimal code implementations, often through comments in the source code Potdar and Shihab 2014 . Detecting and managing SATD has become increasingly important due to its significant impact on long-term software maintainability Maldonado and Shihab 2015 . Over the past decade, various tools and approaches have been proposed to automate the detection of SATD, reflecting a growing recognition of its importance to software maintainability. However, despite theoretical advancements, the transition from proposed solutions to widely accessible, practically implementable tools remains a significant challenge in the field. Several studies have developed automated approaches to identify SATD in source code comments, with tools such as SATDBailiff AlOmar et al. 2022 and DebtHunter Sala et al. 2021 enabling more effective tracking and management of SATD instances. These tools are critical for managing technical debt, as SATD is prevalent in software projects, affecting 2.4% to 31% of files. SATD can persist for extended periods, with a median lifespan ranging from 18 to 172 days and, in some cases, surviving for over 1,000 commits Maldonado et al. 2017 . The availability of reliable tools facilitates better identification and management of SATD, helping developers address technical debt and improve overall software quality. This paper presents a state-of-practice report on the current landscape of SATD detection tools. Based on a recently completed systematic literature review Sutoyo and Capiluppi 2024 , it provides an evaluation of the available software tools, assessing their functionality and limitations across various dimensions, including accessibility, platform compatibility, and performance in real-world applications. In addition to examining the practical aspects of these tools, this paper identifies recurring anti-patterns—such as poor maintenance, lack of interoperability with modern platforms, and inadequate documentation—that hinder their broader adoption and effectiveness. Based on this analysis, we propose actionable corrections to these anti-patterns to improve the sustainability, usability, and future-proofing of SATD detection tools. Our findings offer insights into the current gaps in the state-of-practice and suggest practical improvements that can drive more robust and reliable SATD management in software development. This paper is articulated as follows: Section 2 presents the related work and reviews existing research on SATD detection tools, focusing on their development, adoption, and challenges. Section 3 summarizes the methodology and categorises the tools based on their functionality and availability. Section 4 identifies common anti-patterns in SATD tools availability, while Section 5 proposes actionable strategies for combating these anti-patterns. Finally, Section 6 concludes the paper. 2 Related work Research software plays a crucial role in modern science, but its unavailability or malfunction can have serious repercussions. Lack of access to software and data hinders replication of studies, wastes resources on reinventing existing tools, and limits research opportunities Graham Lee et al. 2021 . Many published papers fail to provide accessible data or documentation on outlier handling, impeding reproducibility Henrik Larsson et al. 2014 . At the same time, insufficient software engineering practices in research can undermine Findable, Accessible, Interoperable, and Reusable (FAIR) principles Graham Lee et al. 2021 , and the absence of source code compromises peer review and may bias subsequent work J. Brito et al. 2020 . To address these issues, experts have long recommended adopting reproducible research practices, which involve publishing both papers and their computational environments L. Madeyski and B. Kitchenham 2017 ; Peng 2011 . This approach can serve as a minimum standard for evaluating scientific claims when full independent replication is not feasible, ultimately enhancing the reliability and transparency of computational research Peng 2011 . Researchers have shared tools to identify SATD to facilitate empirical studies and improve software maintenance. These tools, such as SATDBailiff AlOmar et al. 2022 and SATD Detector Zhongxin Liu et al. 2018 , use text mining and machine learning techniques to automatically detect SATD in source code comments, commit messages, pull requests, and issue trackers Li et al. 2023 . Studies have shown that SATD is common in software projects, affecting 2.4%-31% of files Potdar and Shihab 2014 , and can be effectively identified using automated approaches with high precision and recall Huang et al. 2017 ; Zampetti et al. 2017 . Researchers have also explored specific types of SATD, such as “on-hold” SATD Maipradit et al. 2020 , and investigated the gap between admitted and measured technical debt Pavlič et al. 2022 . These tools and studies contribute to better understanding and managing SATD, improving software quality and maintenance practices. 3 Methodology To evaluate the current state-of-practice of software tools designed to detect SATD, we conducted a systematic literature review (SLR) of academic literature and available tools Sutoyo and Capiluppi 2024 . The review employed search terms such as: (“self-admitted technical debt” OR SATD) OR (“technical debt” AND NLP) AND (detect* OR identif* OR predict*) AND (“software engineering” OR “software development”) . We intentionally used a broad search string without specifically including the term “ tool .” This approach was intentional to capture as many SATD detection approaches as possible, ensuring a comprehensive review and minimizing the risk of overlooking relevant studies. This study builds upon the SLR that identified 68 papers on SATD detection approaches. While the SLR provided a detailed analysis of these approaches, the focus of this paper is narrower, centering specifically on tools identified through the review. We analyze their accessibility and practical utility, highlighting gaps and proposing actionable solutions to address identified anti-patterns. By presenting these contributions, this paper complements the SLR by emphasizing practical insights that can guide the development and improvement of SATD detection tools. From these 68 studies, we carefully reviewed each paper to determine which ones not