Abstract
Background: Many progranunmg environments include automated feedback in the form of hints to help novices learn autonomously. Some experimental studies investigated the impact of automated liints in the immediate, performance and learning retention in that context. Automated feedback is also becoming a popular research topic in the context of formal specification languages, but so far no experimental studies have been conducted to assess its impact while learning such languages. Objective: We aim to investigate the impact of different types of automated hints while learning a formal specification language, not only in terms of immediate performance and learning retention, but also in the emotional response of the students. Method: We conducted a simple one-factor randomised experiment in 2 sessions involving 85 BSc students majoring in CSE. In the 1st session students were divided in 1 control group and 3 experimental groups, each receiving a different type of hint while learning to specify simple, requirements with the Alloy formal specification language. To assess the impact of hints on learning retention, in the 2nd session, 1 week later, students had no hints while formalising requirements. Before and after each session the students answered a standard self-reporting emotional survey to assess their emotional response to the experiment. Results: Of the 3 types of hints considered, only those pointing to the precise location of an error had a positive impact on the immediate performance and none had significant impact in learning retention. Hint availability also causes a significant impact on the emotional response, but no significant emotional :impact exists once hints are no longer available (i.e. no deprivation effects were detected). Conclusion: Although none of the evaluated hints had an impact on learning retention, learning a formal specification language with an environment that provides hints with precise error locations seems to contribute to a better overall experience without apparent drawbacks. Further studies are needed to investigate if other kind of feedback, namely hints combined with some sort of self explanation prompts, can have a positive impact in learning retention.

Abstract
The R/G approach to the development of interfering programs was initiated by the pioneering work of Cliff Jones (1981) on a relational basis. R/G has been the subject of much research since then, most of it deviating from the original relational set-up. This paper looks at such early work from a historical perspective and shows how it can be approached and extended using state-of-the-art relational algebra. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

Abstract
Humans are good at understanding subjective or vague statements which, however, are hard to express in classical logic. Fuzzy logic is an evolution of classical logic that can cope with vague terms by handling degrees of truth and not just the crisp values true and false. Logic is the formal basis of computing, enabling the formal design of systems supported by tools such as model checkers and theorem provers.This paper shows how a model checker such as Alloy can evolve to handle both classical and fuzzy logic, enabling the specification of high-level quantitative relational models in the fuzzy domain. In particular, the paper showcases how QAlloy-F (a conservative, general-purpose quantitative extension to standard Alloy) can be used to tackle fuzzy problems, namely in the context of validating the design of fuzzy controllers. The evaluation of QAlloy-F against examples taken from various classes of fuzzy case studies shows the approach to be feasible.

Abstract
Modern supercomputers host numerous jobs that compete for shared storage resources, causing I/O interference and performance degradation. Solutions based on software-defined storage (SDS) emerged to address this issue by coordinating the storage environment through the enforcement of QoS policies. However, these often fail to consider the scale of modern HPC infrastructures. In this work, we explore the advantages and shortcomings of state-of-the-art SDS solutions and highlight the scale of current production clusters and their rising trends. Furthermore, we conduct the first experimental study that sheds new insights into the performance and scalability of flat and hierarchical SDS control plane designs. Our results, using the Frontera supercomputer, show that a flat design with a single controller can scale up to 2,500 nodes with an average control cycle latency of 41 ms, while hierarchical designs can handle up to 10,000 nodes with an average latency ranging between 69 and 103 ms.

Abstract
TADA is a unique toolkit designed to foster the use and implementation of approximate distributed agreement primitives. Developed in Java, TADA provides ready-to-use implementations of several approximate agreement algorithms, as well as the tools to enable programmers/researchers to easily implement further protocols: A template that enables new protocol implementations to be created by simply changing specific functions; and high-level abstractions for communication and concurrency control. As an example, the toolkit includes a ready-to-use implementation for clock synchronisation between distributed processes. Further use cases can include sensor input stabilisation and distributed machine learning, or other instances of distributed agreement where network synchrony cannot be assumed, byzantine fault tolerance may be required and a bounded divergence in decision values can be tolerated.

Abstract
We present LAZYFS, a new fault injection tool that simplifies the debugging and reproduction of complex data durability bugs experienced by databases, key-value stores, and other data-centric systems in crashes. Our tool simulates persistence properties of POSIX file systems (e.g., operations ordering and atomicity) and enables users to inject lost and torn write faults with a precise and controlled approach. Further, it provides profiling information about the system's operations flow and persisted data, enabling users to better understand the root cause of errors. We use LAZYFS to study seven important systems: PostgreSQL, etcd, Zookeeper, Redis, LevelDB, PebblesDB, and Lightning Network. Our fault injection campaign shows that LAZYFS automates and facilitates the reproduction of five known bug reports containing manual and complex reproducibility steps. Further, it aids in understanding and reproducing seven ambiguous bugs reported by users. Finally, LAZYFS is used to find eight new bugs, which lead to data loss, corruption, and unavailability.