Large language models (LLMs) have demonstrated the promise to revolutionize the field of software engineering. Among other things, LLM agents are rapidly gaining momentum in software development, with practitioners reporting a multifold increase in productivity after adoption. Yet, empirical evidence is lacking around these claims. In this paper, we estimate the causal effect of adopting a widely popular LLM agent assistant, namely Cursor, on development velocity and software quality. The estimation is enabled by a state-of-the-art difference-in-differences design comparing Cursor-adopting GitHub projects with a matched control group of similar GitHub projects that do not use Cursor. We find that the adoption of Cursor leads to a statistically significant, large, but transient increase in project-level development velocity, along with a substantial and persistent increase in static analysis warnings and code complexity. Further panel generalized-method-of-moments estimation reveals that increases in static analysis warnings and code complexity are major factors driving long-term velocity slowdown. Our study identifies quality assurance as a major bottleneck for early Cursor adopters and calls for it to be a first-class citizen in the design of agentic AI coding tools and AI-driven workflows.

Talks at practitioner-focused open-source software conferences are a valuable source of information for software engineering researchers. They provide a pulse of the community and are valuable source material for grey literature analysis. We curated a dataset of 24,669 talks from 87 open-source conferences between 2010 and 2021. We stored all relevant metadata from these conferences and provide scripts to collect the transcripts. We believe this data is useful for answering many kinds of questions, such as: What are the important/highly discussed topics within practitioner communities? How do practitioners interact? And how do they present themselves to the public? We demonstrate the usefulness of this data by reporting our findings from two small studies: a topic model analysis providing an overview of open-source community dynamics since 2011 and a qualitative analysis of a smaller community-oriented sample within our dataset to gain a better understanding of why contributors leave open-source software.

Feature flags (a.k.a feature toggles) are a mechanism to keep new features hidden behind a boolean option during development. Flags are used for many purposes, such as A/B testing and turning off a feature more easily in case of failures. While software engineering research on feature flags is burgeoning, examples of software projects using flags rarely come from outside commercial and private projects, stifling academic progress. To address this gap, in this paper we present a novel semi-automated mining software repositories approach to detect feature flags in open-source projects, based on analyzing the projects' commit messages and other project characteristics. With our approach, we search over all open-source GitHub projects, finding multiple thousand plausible and active candidate feature flagging projects. We manually validate projects and assemble a dataset of 100 confirmed feature flagging projects. To demonstrate the benefits of our detection technique, we report on an initial analysis of feature flags in the validated sample of 100 projects, investigating practices that correlate with shorter flag lifespans (typically desirable to reduce technical debt), such as using the issue tracker and having a flag owner.

Previous research shows that developers spend most of their time understanding code. Despite the importance of code understandability for maintenance-related activities, an objective measure of it remains an elusive goal. Recently, Scalabrino et al. reported on an experiment with 46 Java developers designed to evaluate metrics for code understandability. The authors collected and analyzed data on more than a hundred features describing the code snippets, the developers' experience, and the developers' performance on a quiz designed to assess understanding. They concluded that none of the metrics considered can individually capture understandability. Expecting that understandability is better captured by a combination of multiple features, we present a reanalysis of the data from the Scalabrino et al. study, in which we use different statistical modeling techniques. Our models suggest that some computed features of code, such as those arising from syntactic structure and documentation, have a small but significant correlation with understandability. Further, we construct a binary classifier of understandability based on various interpretable code features, which has a small amount of discriminating power. Our encouraging results, based on a small data set, suggest that a useful metric of understandability could feasibly be created, but more data is needed.

Continuous Integration (CI) services, which can automatically build, test, and deploy software projects, are an invaluable asset in distributed teams, increasing productivity and helping to maintain code quality. Prior work has shown that CI pipelines can be sophisticated, and choosing and configuring a CI system involves tradeoffs. As CI technology matures, new CI tool offerings arise to meet the distinct wants and needs of software teams, as they negotiate a path through these tradeoffs, depending on their context. In this paper, we begin to uncover these nuances, and tell the story of open-source projects falling out of love with Travis, the earliest and most popular cloud-based CI system. Using logistic regression, we quantify the effects that open-source community factors and project technical factors have on the rate of Travis abandonment. We find that increased build complexity reduces the chances of abandonment, that larger projects abandon at higher rates, and that a project's dominant language has significant but varying effects. Finally, we find the surprising result that metrics of configuration attempts and knowledge dispersion in the project do not affect the rate of abandonment.

The Android platform is designed to support mutually untrusted third-party apps, which run as isolated processes but may interact via platform-controlled mechanisms, called Intents. Interactions among third-party apps are intended and can contribute to a rich user experience, for example, the ability to share pictures from one app with another. The Android platform presents an interesting point in a design space of module systems that is biased toward isolation, extensibility, and untrusted contributions. The Intent mechanism essentially provides message channels among modules, in which the set of message types is extensible. However, the module system has design limitations including the lack of consistent mechanisms to document message types, very limited checking that a message conforms to its specifications, the inability to explicitly declare dependencies on other modules, and the lack of checks for backward compatibility as message types evolve over time. In order to understand the degree to which these design limitations result in real issues, we studied a broad corpus of apps and cross-validated our results against app documentation and Android support forums. Our findings suggest that design limitations do indeed cause development problems. Based on our results, we outline further research questions and propose possible mitigation strategies.