Search

International audience; ProbZelus is a synchronous probabilistic language for the design of reactive probabilistic models in interaction with an environment. Reactive inference methods continuously learn distributions over the unobserved parameters of the model from statistical observations. Unfortunately, this inference problem is in general intractable. Monte Carlo inference techniques thus rely on many independent executions to compute accurate approximations. These methods are expensive but can be parallelized. We propose to use JAX to parallelize ProbZelus reactive inference engine. JAX is a recent library to compile Python code which can then be executed on massively parallel architectures such as GPUs or TPUs. In this paper, we describe a new reactive inference engine implemented in JAX and the new associated JAX backend for ProbZelus. We show on existing benchmarks that our new parallel implementation outperforms the original sequential implementation for a high number of particles.

Conventionally, callbacks and inversion of control have been the main tools to structure event-driven applications. Sadly, those patterns constitute a well-known source of design problems. The Reactive Programming (RP) paradigm has arisen as an approach to mitigate these problems. Yet, little evidence has been provided regarding the advantages of RP, and concerns have also arisen about the API usability of RP libraries given their disparate number of operators. In this work, we conduct a study on GitHub (GH) and Stack Overflow (SO) and explore three Reactive Extensions (Rx) libraries (RxJava, RxJS, and RxSwift) with the most GH projects to understand how much the vast Rx operators are being used. Also, we examine Rx SO posts to complement the results from the GH exploration by understanding the problems faced by RP developers and how they relate with the operators' frequencies found in open source projects. Results reveal that, in spite of its API size, the great majority of the Rx operators are actually being used (95.2%), with only a few, mostly related to RxJava, not being utilized. Also, we unveil 23 topics from SO with more posts concerning the Stream Abstraction (36.4%). Posts related to Dependency Management, Introductory Questions, and iOS Development figure as relevant topics to the community. The findings herein present can not only stimulate advancements in the field by understanding the usage of RP API and the main problems faced by developers, but also help newcomers in identifying the most important operators and the areas that are the most likely to be relevant for a RP application.

This paper presents Poker, a visual instrumentation platform for reactive programs. Similar to other platforms, Poker features a visual dashboard that allows the programmer to in- spect the flow of values through the reactive program. The novelty of Poker is that: (a) It features a canvas of so-called probes that can be dynamically wired into a running reactive program in order to instrument the running system. (b) In addition to focusing on the values flowing through the pro- gram, a probe can measure a particular property about the way these values flow through the instrumented program. (c) The set of probes is open because a probe is programmed in the same language as the instrumented program. Poker is implemented for Stella, an experimental reactive programming language. The paper uses an application written in Stella to motivate the concepts provided by Poker. We show 4 different probes that help us understand the behaviour of the application and we measure the overhead of using Poker on the running application with some preliminary benchmarks.

International audience; Modern services running in cloud and edge environments need to be resource-efficient to increase deployment density and reduce operating costs. Asynchronous I/O combined with asynchronous programming provides a solid technical foundation to reach these goals. Reactive programming and reactive streams are gaining traction in the Java ecosystem. However, reactive streams implementations tend to be complex to work with and maintain. This paper discusses the performance of the three major reactive streams compliant libraries used in Java applications: RxJava, Project Reactor, and SmallRye Mutiny. As we will show, advanced optimization techniques such as operator fusion do not yield better performance on realistic I/O-bound workloads, and they significantly increase development and maintenance costs.

Sorting is a central problem in computer science and one of the key components of many applications. To the best of our knowledge, no reactive programming implementation of sorting algorithms has ever been presented. In this paper we present reactive implementations of so-called sorting networks. Sorting networks are networks of comparators that are wired up in a particular order. Data enters a sorting network along various input wires and leaves the sorting network on the same number of output wires that carry the data in sorted order. This paper shows how sorting networks can be expressed elegantly in a reactive programming language by aligning the visual representation of a sorting network with the canonical DAG representation of reactive programs. We use our own experimental language called Haai to do so. With a limited number of built-in higher-order reactive programs, we are able to express sorting networks for bubble sort, insertion sort, bitonic sort, pairwise sort and odd-even merge sort.

Many industrial IoT applications process sensor data over distributed networks to monitor devices in real-time. Since the sensor telemetries are transmitted over networks as events, imperative and event-driven programs are often used to handle IoT data. However, the inverted control flow and shared global states of these imperative programs make them difficult to interface with synchronized computation on IoT data. This problem is further complicated for high-frequency data such as electric signals, which may require dynamic adjustment to data sampling rate to operate under the constraints of network and system. In this paper, we propose a push-pull reactive programming model for IoT application to address this challenge. This model uses push-streams for asynchronous computation such as data capturing and user controls and uses pull streams for synchronized computation such as data analysis. This model is simpler than push-based models by avoiding the complexity of glitch prevention through re-sampling in pull-streams. It is also more flexible than pull-based models by allowing dynamic adjustment of the sampling rate to maintain real-time speed of the IoT computation. The push-stream has a monadic interface, which converts to a pull stream through buffering. A pull stream converts to a push-stream when driven by a clock. The dynamic switching of our streams is based on a monadic abstraction called AsyncM that uses continuation passing style and a form of cancellation token for asynchronous control. Our model is simple and can use threads or event callbacks for concurrency.

RxJS is a popular library to implement data-flow-oriented applications with JavaScript using reactive programming principles. This way of programming bears new challenges for traditional debuggers: Their focus on imperative programming limits their applicability to problems originated in the declarative programming paradigm. The goals of this paper are: (i) to understand how software engineers debug RxJS-based applications, what tools do they use, what techniques they apply; (ii) to understand what are the most prevalent challenges they face while doing so; and (iii) to provide a course of action to resolve these challenges in a future iteration on the topic. We learned about the debugging habits of ten professionals using interviews, and hands-on war story reports. Based on this data, we designed and executed an observational study with four subjects to verify that engineers predominantly augment source code with manual trace logs instead of using specialized debugging utilities. In the end, we identified the lack of fully integrated RxJS-specific debugging solutions in existing development environments as the most significant reason why engineers do not make use of such tools. We decided to elaborate on how to resolve this situation in our future work.

Different activation mechanisms for Context-Oriented Programming (COP) like implicit activations have been proposed, increasing COP opportunities to be applied in real scenarios. However, activation mechanisms and base code definitions are insufficiently decoupled, as conditionals to activate layers require base code variable references. This hinders reuse, evolution, and modular reasoning of COP and base code, and therefore, uses of COP in real scenarios. This paper proposes interfaces, which are shared abstractions to communicate activation mechanisms and base code in a decoupled manner. Using these interfaces, an object can exhibit its internal state and behaviors, and conditionals use them to (de)activate layers. As layers are planned to be (re)used in different applications, developers can use interfaces to overcome the incompatibility between values exposed by a particular base code and values required by a layer. In addition, as a layer is a plain object, it can use an interface to exhibit the conditional evaluation of its activation to other layers to resolve conflicts among activations of layers. We apply this proposal to implicit activations in which evaluations of conditionals implicitly (de)activate layers. Finally, we illustrate the benefits of this proposal through RI-JS, a practical JavaScript library that currently supports interfaces, reactive activations (implementation variant for implicit activations), global and dynamic deployment, enter and exit transition processes, and partial methods.

Parallelization of static analyses is necessary to scale to real-world programs, but it is a complex and difficult task and, therefore, often only done manually for selected high-profile analyses. In this paper, we propose a programming model for semi-implicit parallelization of static analyses which is inspired by reactive programming. Reusing the domain-expert knowledge on how to parallelize anal- yses encoded in the programming framework, developers do not need to think about parallelization and concurrency issues on their own. The programming model supports stateful computations, only requires monotonic computations over lattices, and is independent of specific analyses. Our evaluation shows the applicability of the programming model to different analyses and the importance of user-selected scheduling strategies. We implemented an IFDS solver that was able to outperform a state-of-the-art, specialized parallel IFDS solver both in absolute performance and scalability.

In user interfaces on Web and desktop applications, the system must quickly respond to user inputs and accurately display the result. Current solutions for user interfaces often respond too slowly and display temporarily inaccurate or misleading states, resulting in low user satisfaction. To alleviate this problem, we develop a reactive programming approach that encodes the complex relationships among the user interface components in a dependency graph and then uses the graph to order the updates of the components without violating the dependency constraints. This enables more timely updates and more accurate visualizations, potentially providing users with a more satisfying experience. We evaluate our approach by comparing its performance with important alternative reactive libraries for user interfaces.

Reactive programming over a network is a challenging task because efficient elimination of temporary violations of data flow invariants, known as glitches, in a distributed setting is still an open issue. In this paper, we propose a method for constructing a distributed reactive programming system of which runtime guarantees the properties of single source glitch-freedom and the robustness against out-of-order messages. Based on the method, we developed a purely functional reactive programming language XFRP whose compiler produces Erlang code. Using some examples, we show that the proposed method is beneficial for constructing distributed reactive applications without suffering from inconsistencies.

This paper presents the design of a simple context-oriented programming extension to Emfrp, a purely functional reactive programming language for resource-constrained embedded systems. The language provides declarative and straightforward ways to describe various reactive behaviors in embedded systems. However, because of the static nature of the language, Emfrp lacks suitable capabilities to modularize adaptive behaviors. The proposed extension introduces a simple layer mechanism with implicit layer activation. It also provides a feature to describe responses to events that correspond to the moments of layer activations and deactivations. The extension can improve the modularity of Emfrp programs by separating the descriptions of context-dependent behaviors, and hence eliminating various cross-cutting code fragments.

As reactive systems, such as cyber-physical systems and the Internet of Things, become increasingly important, time-varying values, also known as signals, are playing an important role in software development. Although reactive systems require the change histories of some signals to be stored for various purposes such as post analysis and simulation, current programming languages do not provide a way to declare that signals are persistent. This paper proposes a method that realizes persistent signals in a reactive programming language, where (1) every update to each persistent signal is recorded in a time-series database, which can be seen as a part of the programming language runtime; and (2) persistent signals support a convenient time-oriented query mechanism. In this approach, each signal in the reactive programming language is seamlessly connected with the time-series database. This method is implemented as an extension of SignalJ, a Java-based reactive programming language that supports signals. In the implementation, the persistent signal mechanism is integrated with TimescaleDB, a PostgreSQL-based time-series database. In preliminary performance evaluations, our implementation had good responsiveness on most tests, indicating its feasibility for use in many applications.

Concurrency control is difficult in JavaScript programs, where event race due to asynchronous computation is a major source of errors. While methods such as promises, cancellation tokens, and reactive programming offer their own advantages in addressing this problem, none offer a complete solution. In this work, we present an integrated solution for concurrency control of JavaScript using a library, arrowjs, which is based on the abstraction of arrows. Arrowjs uses continuation passing style to chain callbacks and it implicitly generates progress objects to manage concurrency. Arrowjs can implement a form of push-based reactive programming, where event streams are arrow loops communicating through shared memory. Arrowjs thus provides interoperability between thread-like callback chains and event streams with a uniform concurrency control mechanism.

Functional Reactive Programming is an approach to declarative programming of reactive systems by describing interactions between time-varying values. FRP implementations are often realised as an embedding in a functional host language, making for very expressive reactive programming frameworks. However, this expressiveness comes at a cost: current embedded FRP implementations incur substantial performance overheads, in particular for values that (notionally) vary continuously. The basic idea of FRP is closely related to synchronous data-flow and continuous system simulation languages. In contrast to FRP, these handle values that vary continuously efficiently, but are less expressive. This paper seeks to bridge this gap by proposing a novel approach to embedded FRP-implementation that uses the fundamental implementation approach of synchronous datalow and simulation languages for efficient handling of continuously varying values, while retaining the expressiveness normally associated with FRP, as well as paying attention to values that only change relatively infrequently. These ideas are applicable beyond FRP, for example for implementing flexible embedded simulation languages. We evaluate our approach on a range of benchmarks, including an existing full-fledged video game where using our new FRP implementation as a drop-in replacement for the old one gave a three-fold performance improvement.

Log analysis is required in many domains, and especially in the emerging field of cloud computing. Cloud applications are often built by composing diverse services. When something goes wrong, finding the root cause of the problem can be difficult. Many services are only reachable through their Application Programming Interfaces (APIs) with no possibility for live introspection. In this context, logs become an essential tool for monitoring and debugging. Cloud services typically generate very large quantities of log messages, with formats that may not be well specified and may vary over time. In this paper, we present CloudLens, a language for the analysis of semi-structured textual data as found in logs, and specify its formal semantics. CloudLens is a reactive language and views logs as streams of objects. Our objective is to facilitate exploring the contents of logs interactively and to write reusable analyses succinctly, using familiar constructs. We implemented an interpreter for the Apache Zeppelin notebook to provide an interactive IDE. Our prototype implementation is open source and we report on a detailed case study using logs from the Apache OpenWhisk project.

Chatbots are reactive applications with a conversational interface. They are usually implemented as compositions of client-side components and cloud-hosted services, including artificial-intelligence technology. Unfortunately, programming such reactive multi-tier applications with traditional programming languages is cumbersome. This paper introduces wcs-ocaml, a new multi-tier chatbot generator library designed for use with the reactive language ReactiveML. The paper explains our library with small didactic examples throughout, and closes with a larger case-study of a chatbot for authoring event-processing rules.

Participatory sensing (PS) platforms enable stakeholders to collect, analyse and visualise data for a particular interest. Despite high societal demand, developing a new PS platform remains a labour-intensive, nonreusable process that requires high technical expertise. We present DISCOPAR, a visual reactive flow-based domain-specific language geared towards the construction of reusable citizen observatories. With DISCOPAR, users interact with visual components to implement the various elements of a PS platforms without having to worry about its underlying technological complexities. We validate our approach through experiments using real-world empirical usability studies of ICT-agnostic users. The results show that DISCOPAR allows users with limited technological knowledge to create their own PS platform.

A live programming environment allows programmers to edit programs while they are running. This means that successive "edit steps" must not allow a programmer to bring the program in a form that does not make any sense to the underlying language processor (i.e., parser, compiler,...). Many live programming environments therefore rely on disciplined edit steps that are based on language elements such as objects, classes, and methods. Textual modifications to these elements are not seen as edit steps until some "accept" button is hit. Unfortunately, no such elements exist in current reactive languages. We present a new reactive language, called Haai, that is based on first-class higher-order reactors. Linguistically, Haai programs correspond to reactors or compositions of reactors. At run-time, reactors produce an infinite stream of values just like signals and behaviours in existing languages. Haai's live programming environment relies on textual modifications of entire reactors as its basic edit steps. Changing a reactor automatically updates all occurrences of that reactor in the reactive program, while it is running.

Processing data at different rates is generally a hard problem in reactive programming. Buffering problems, lags, and concurrency issues often occur. Many of these problems are clock errors, where data at different rates is combined incorrectly. Techniques to avoid clock errors, such as type-level clocks and deterministic scheduling, exist in the field of synchronous programming, but are not implemented in general-purpose languages like Haskell. Rhine is a clock-safe library for synchronous and asynchronous Functional Reactive Programming (FRP). It separates the aspects of clocking, scheduling and resampling from each other, and ensures clock-safety at the type level. Concurrent communication is encapsulated safely. Diverse reactive subsystems can be combined in a coherent, declarative data-flow framework, while correct interoperability of data at different rates is guaranteed by type-level clocks. This provides a general-purpose framework that simplifies multi-rate FRP systems and can be used for game development, media applications, GUIs and embedded systems, through a flexible API with many reusable components.

In recent years, there has been tremendous progress in automated synthesis techniques that are able to automatically generate code based on some intent expressed by the programmer. A major challenge for the adoption of synthesis remains in having the programmer communicate their intent. When the expressed intent is coarse-grained (for example, restriction on the expected type of an expression), the synthesizer often produces a long list of results for the programmer to choose from, shifting the heavy-lifting to the user. An alternative approach, successfully used in end-user synthesis is programming by example (PBE), where the user leverages examples to interactively and iteratively refine the intent. However, using only examples is not expressive enough for programmers, who can observe the generated program and refine the intent by directly relating to parts of the generated program. We present a novel approach to interacting with a synthesizer using a granular interaction model. Our approach employs a rich interaction model where (i) the synthesizer decorates a candidate program with debug information that assists in understanding the program and identifying good or bad parts, and (ii) the user is allowed to provide feedback not only on the expected output of a program, but also on the underlying program itself. That is, when the user identifies a program as (partially) correct or incorrect, they can also explicitly indicate the good or bad parts, to allow the synthesizer to accept or discard parts of the program instead of discarding the program as a whole. We show the value of our approach in a controlled user study. Our study shows that participants have strong preference to using granular feedback instead of examples, and are able to provide granular feedback much faster.

Reactive Programming is a style of programming that provides developers with a set of abstractions that facilitate event handling and stream processing. Traditional debug tools lack support for Reactive Programming, leading developers to fallback to the most rudimentary debug tool available: logging to the console. In this paper, we present the design and implementation of RxFiddle, a visualization and debugging tool targeted to Rx, the most popular form of Reactive Programming. RxFiddle visualizes the dependencies and structure of the data flow, as well as the data inside the flow. We evaluate RxFiddle with an experiment involving 111 developers. The results show that RxFiddle can help developers finish debugging tasks faster than with traditional debugging tools.

The Internet of things (IoT) has expanded at a fantastic pace in the last years. It will be only a matter of years before every single daily object will contain embedded sensors and actuators connected through the Internet, producing massive amounts of data. Meanwhile, efforts are being carried out to enable IoT objects to be accessible through the Web as well. Thanks to those efforts, a new class of graphical tools, namely physical mashups, that explores data flow programming to connect Web components and Internet of things' devices easily, can now be used to build applications that mix the physical world with the virtual one. In parallel, the concept of Complex Event processing has emerged as a response to the increase of data from different types of distributed applications due to the need for (quasi) real-time processing of data. In this paper we propose the combination of both approaches, CEP and WoT mashups, to build a simple approach to tackle this new avalanche of data in the context of IoT applications. We apply Reactive Programming to implement the CEP operators that we provide as extensions of an existing WoT platform. A use case scenario is presented to demonstrate the approach.

Reactive programming is a recent programming paradigm that specifically targets reactive applications. Over the years, a number of reactive languages have been proposed, with different combinations of features, and various target domains. Unfortunately, there is a lack of knowledge about the experience of developing software applications with reactive languages. As a result, a number of design choices in reactive programming languages remain disconnected from experience and the applicability of reactive programming to various domains remains unclear. To bridge this gap, we report on our experience of developing reactive applications as well as teaching reactive programming in REScala, which we collected over several years of research and practice.

Cyber-Physical Systems (CPS) are comprised of a network of devices that vary widely in complexity, ranging from simple sensors to autonomous robots. Traditionally, controlling and sensing these devices happens through API communication, in either push or pull-based fashion. We argue that the computational power of these devices is converging to the point where they can do autonomous computations. This allows application programmers to run programs locally on the sensors, thereby reducing the communication and workload of more central command and control entities. This work introduces the Potato framework that aims to make programming CPS systems intuitively easy and fast. Potato is based on three essential mechanisms: failure handling by means of leasing, distribution by means of first-class reactive programs, and intentional retroactive designation of the network by means of capabilities and dynamic properties. In this paper we focus on the reactive capabilities of our framework. Potato enables programmers to create and deploy first-class reactive programs on CPS devices at run time, abstracting away from the API approach. Each node in the network is equipped with a minimal actor-based middleware that can execute first-class reactive programs. We have implemented Potato as a library in Elixir and have used it to implement several small examples.

In his famous paper entitled "Tackling the Awkward Squad", Peyton Jones studies how features that traditionally did not fit in the functional programming paradigm can be added to a functional language via careful language design (e.g. using monads), instead of allowing programmers to sprinkle around impure expressions and ad-hoc library calls, thereby turning the entire program into a non-functional program. Similarly, in this paper, we identify a number of code characteristics that do not map onto the reactive programming paradigm but that are present in many real life reactive programs. We propose a novel Actor-Reactor model that can serve as the basis for future language designs that allow a programmer to use the awkward squad without making the reactive parts of the program accidentally non-reactive.

Distributed reactive programming enables programmers to reuse the abstractions provided by reactive programming to elegantly implement distributed systems. However, distributed reactive approaches have thus far neglected to address an inherent property of distributed systems: partial failures. This forces programmers to either disregard failures and write poor distributed code or try to detect failures manually (e.g. through time-outs and heartbeats). Moreover, this prohibits distributed reactive runtimes to garbage collect remote references to failed parts of the reactive network. In this paper we present a first attempt at failure handling for distributed reactive applications. To this end we introduce the novel concept of leased signals which allow both programmer and runtime to react to partial failures in distributed reactive applications. We implement leased signals in a distributed reactive TypeScript framework for the development of microservice applications.

To achieve scalability and flexibility in larger applications a new approach arises, named by Microservices (MS). However MS architectures are at their inception and are even more a concept than a fully mature design pattern. One of the hardest topics in this approach is how to properly migrate or develop a single microservice, in terms of scope, efficiency and dependability. In this sense, this work proposes a new architectural model based on high-level architecture pattern of reactive programming to the internal structure of a new microservice. The new model of microservices are internally coordinated by asynchronous queues, which allowed to preserve compatibility with most monolithic components and provide an encapsulation process to enable its continuity. A comparative study between the standard approach and the proposed architecture was carried out to measure the eventual performance improvement of the new strategy.

Many programs have crosscutting concerns for which neither procedural nor object-oriented programming adequately modularize, which has led to the idea of aspect-oriented programming (AOP). However, AOP has not found favor with the programming languages community due to a belief that AOP breaks classical modularity and modular reasoning. We propose a new AOP programming model that enables both crosscutting modularity and modular reasoning. This model is implemented by AspectJML, a general-purpose aspect-oriented extension to Java. It supports modular crosscutting concerns using key object-oriented mechanisms, such as hierarchical structure, and allows reasoning that scales to ever-larger programs.

Functional Reactive Programming (FRP) allows interactive applications to be modelled in a declarative manner using time-varying values. For practical reasons, however, operational constraints are often imposed, such as having a fixed time domain, time always flowing forward, and limiting the exploration of the past. In this paper we show how these constraints can be overcome, giving local control over the time domain, the direction of time and the sampling step. We study the behaviour of FRP expressions when time flows backwards, and demonstrate how to synchronize subsystems running asynchronously and at different sampling rates. We have verified the practicality of our approach with two non-trivial games in which time control is central to the gameplay.

We present a vision of genetic improvement firmly embedded in, and supported by, programming languages. Genetic improvement has already been envisioned as the next compiler, which would take human written programs as input and return versions optimised with respect to various objectives. As an intermediate stage, or perhaps to complement the fully automated vision, we imagine genetic improvement processes that are hinted at and directed by humans but understood and undertaken by programming languages and their runtimes, via interactions through the source code. We examine existing similar ideas and examine the benefits of embedding them within programming languages.

As computer programming becomes more important to various fields and disciplines and as it is more commonly taught in education settings, the number of end-users with basic programming experience is increasing. The importance of being able to easily and quickly develop programs has prompted research in "opportunistic' programming methods. This research contributes to this domain by introducing a platform called Intuito, designed for programming physical components (sensors and actuators) through direct "programming by demonstration' techniques. Our approach is to offer users a tangible system that maps the user's actions with sensors and actuators into editable text-based code by inferring the user's intentions. We present our initial hardware and software prototype with an in-lab preliminary evaluation of this system.

A recent trend in application development for the web is a move towards rich internet applications (RIAs). As more and more of the application logic is moved to the client, RIAs can benefit from concurrency in order to increase overall performance as well as responsiveness of the application. Additionally, RIAs are often written in an event-driven style of programming to react to incoming events of a multitude of services that are integrated within the application. In this paper we argue that, while individual technologies exist to tackle both concerns, these technologies cannot easily be integrated in an ad hoc way. To increase the modularisability and composability of RIAs we propose a new programming model based on actors and reactors that encapsulate different parts of the application. We show that our model is able to exploit some of the available concurrency while reducing the required amount of event-driven code. Both actors and reactors are modular components that can be glued together via a unified communication mechanism. We evaluate our approach by means of a motivating example of a collaborative code editor.

With ubiquitous computing, autonomous cars, and cyber-physical systems (CPS), adaptive software becomes more and more important as computing is increasingly context-dependent. Role-based programming has been proposed to enable adaptive software design without the problem of scattering the context-dependent code. Adaptation is achieved by having objects play roles during runtime. With every role, the object's behavior is modified to adapt to the given context. In recent years, many role-based programming languages have been developed. While they greatly differ in the set of supported features, they all incur in large runtime overheads, resulting in inferior performance. The increased variability and expressiveness of the programming languages have a direct impact on the run-time and memory consumption. In this paper we provide a detailed analysis of state-of-the-art role-based programming languages, with emphasis on performance bottlenecks. We also provide insight on how to overcome these problems.

Classically, computers have been used as knowledge discovery tools insofar as the result of executing a program provides useful insight. For instance, the solution of a differential equation may help us understand the natural world, the value of a parameter of a statistical model may help us understand the probabilistic structure of a domain, the variable assignment maximising an objective function may help to further business goals. A secondary class of knowledge discovery stems from the act of using a programming language. By modeling a domain computationally, the developer can discover new and interesting properties of that domain, and better convey those insights to others. The purpose of this work is twofold: First, we want to show that programming languages can help their users achieve knowledge discovery moments and, secondly, that this property is the least exploited feature of programming languages in the general science community. We want to outline a research program with the objective of making scientific programming more efficient in its ultimate goal of knowledge discovery.

This paper introduces NetsBlox, a visual programming environment for learning distributed programming principles. Extending both the visual formalism and open source code base of Snap!, NetsBlox provides two accessible distributed programming abstractions to simplify the process of creating networked applications: message passing and Remote Procedure Calls (RPC). Messaging passing allows NetsBlox applications to send data to other connected NetsBlox clients. Remote Procedure Calls enable seamless integration of third party services, such as Google Maps, weather, traffic and other public domain data sources, into NetsBlox applications. Other RPCs help coordinating distributed clients which may be difficult for novice programmers allowing the user to more quickly create captivating and sophisticated applications. These abstractions empower users to develop networked programs, including multi-player games and client-server applications. By providing networking support, NetsBlox not only allows users to learn distribute programming concepts but also makes programming more engaging by incorporating diverse services available on the web.

Blocks-based, graphical programming environments are increasingly becoming the way that novices are being introduced to the practice of programming and the field of computer science more broadly. An open question surrounding the use of such tools is how well they prepare learners for using more conventional text-based programming languages. In an effort to address this transition, new programming environments are providing support for both blocks-based and text-based programming. In this paper, we present findings from a study investigating how learners use a dual-modality environment where they can choose to work in either a blocks-based or text-based interface, moving between them as they choose. Our analysis investigates what modality learners choose to work in, and if and why they move from one representation to the other within a single project. We conclude with a discussion of design implications and future directions for this work. This work contributes to our understanding of the affordances of blocks-based programming environments and advances our knowledge on how best to utilize them.

Programming mobile manipulator robots requires specialized knowledge of robot perception, manipulation, and software frameworks. This puts robot programming out of reach for programmers who have not invested the time to learn these skills. The goal of our research is to make it possible for all programmers to develop applications for mobile manipulators such as clearing a table, fetching items, or playing games with people. This extended abstract describes our current work on Code3, a framework that all programmers can use to program robots. A user study we conducted showed that non-roboticist programmers were able to use our system to program a PR2 robot to clear a table and retrieve snacks. We also discuss our plans to extend this system to make it more useful and usable for real-world use cases.

This paper introduces Code3, a system for user-friendly, rapid programming of mobile manipulator robots. The system is designed to let non-roboticists and roboticists alike program end-to-end manipulation tasks. To accomplish this, Code3 provides three integrated components for perception, manipulation, and high-level programming. The perception component helps users define a library of object and scene parts that the robot can later detect. The manipulation component lets users define actions for manipulating objects or scene parts through programming by demonstration. Finally, the high-level programming component provides a drag-and-drop interface with which users can program the logic and control flow to accomplish a task using their previously specified perception and manipulation capabilities. We present findings from an observational user study with non-roboticist programmers (N=10) that demonstrate their ability to quickly learn Code3 and program a PR2 robot to do manipulation tasks. We also demonstrate how the system is expressive enough for an expert to rapidly program highly complex manipulation tasks like playing tic-tac-toe and reconfiguring an object to be graspable.

Nodes with multiple GPUs are becoming the platform of choice for high-performance computing. However, most applications are written using bulk-synchronous programming models, which may not be optimal for irregular algorithms that benefit from low-latency, asynchronous communication. This paper proposes constructs for asynchronous multi-GPU programming, and describes their implementation in a thin runtime environment called Groute. Groute also implements common collective operations and distributed work-lists, enabling the development of irregular applications without substantial programming effort. We demonstrate that this approach achieves state-of-the-art performance and exhibits strong scaling for a suite of irregular applications on 8-GPU and heterogeneous systems, yielding over 7x speedup for some algorithms.

Developing modern collaborative applications burdens the programmer with local event handling (e.g. user interaction), remote event handling (e.g. updates from the server) and shared state (e.g. in order to allow operations while being disconnected). Several solutions have been developed at the programming language level in order to reduce the complexity of these aspects. On one hand, distributed reactive models (e.g. DREAM) tackle both local and remote event handling. On the other hand recent replicated consistency models (e.g. CRDT's and CloudTypes) hide the complexity of shared, replicated state. Both solutions only partially alleviate the complexity associated with developing collaborative applications. To the best of our knowledge, none or very little effort has been undertaken to provide a single unified model able to tackle both event handling and shared state. In this paper we argue the need for such a united model. To that end we present Direst, a domain specific language which enhances traditional reactive abstractions (i.e. signals) with replication and consistency features. Direst reduces the complexities of writing truly collaborative applications by providing a framework in which elegantly handling events and easily managing shared state are not mutually exclusive.

End-user programming, a frequently recurring dream, has thus far eluded large-scale, complex applications. Very real, hard questions stand in the way of its realization. How can its languages and tools support: (1) The development of applications with large data sets and sophisticated computation? (2) The co-development by end-users and professional developers when the complexity of an application demands it? (3) Beyond development, the maintenance, distribution, monitoring, and integration with other applications and services? We discuss our approach to these questions, as implemented in the LogicBlox Modeler. We discuss its use in developing applications for governments, major financial institutions, and large global retailers. We highlight the essential synergies between Programming Languages, Software Engineering, and Database research to achieve self-service at scale, and present open questions to which we look to the FSE community for inspirations and solutions.

The current event-based model of web client programming lacks of a high level abstraction for concurrency and communication when many interactions are involved. The design of JavaScript runtime in the browser is very simple and chains steps of input handling and computation of output in a sequential way. This definition describes exactly a subset of programs well handled by the synchronous-reactive model. It proposes constructs to express parallel tasks communicating through broadcasted signals enforcing a static hypothesis of determinism, coherency and causality that improve programs composition. It is then interesting to consider client events as inputs and web view elements as outputs of a synchronous-reactive program. We describe here the design of pendulum, a language extension implementing those principles and targeting web client programming, which generates fast sequential code.

Over the last years, a number of languages based on data flow abstractions have been proposed in different important areas including Big Data, stream processing, reactive programming, real time analytics. While there is a general agreement that the data flow style simplifies the access to such complex systems compared to low level imperative APIs, this design has been substantiated by little evidence. In this paper, we advocate a systematic investigation of the design principles of data flow languages and suggest important research questions that urge to be addressed.

We present a straightforward, call-by-value embedding of a small logic programming language with a simple complete search. We construct the entire language in 54 lines of Racket---half of which implement unification. We then layer over it, in 43 lines, a reconstruction of an existing logic programming language, miniKanren, and attest to our implementation's pedagogical value. Evidence suggests our combination of expressiveness, concision, and elegance is compelling: since microKanren's release, it has spawned over 50 embeddings in over two dozen host languages, including Go, Haskell, Prolog and Smalltalk.

The spatial libraries that have been developed by Microsoft, IBM and Oracle have substantially changed the capabilities of geospatial computing. These libraries implement several functionalities that include intersection, distance, and area for various geospatial objects. These libraries came out to address a wealth of use cases that were challenging in that era. As time goes by, GPS devices and location-aware mobile technologies increased the demand for geospatial computing, in general, and for real time geostreaming, in particular. Existing commercial spatial libraries were originally designed to support operations on stationary objects with limited or no capabilities for moving objects. In this paper, we introduce the RxSpatial library, a real time reactive spatial library for spatiotemporal stream query processing. RxSpatial provides, (1) a front-end, which is a programming interface for developers who are familiar with the Microsoft. NET Reactive framework and the Microsoft SQL Server Spatial Library, and (2) a back-end for processing spatial operations in a streaming fashion. RxSpatial provides the programming convenience at the front end and the query processing efficiency at the back end.

DesignScript is a multi-paradigm domain-specific end-user language and modelling environment for architectural and engineering computation. DesignScript implements both visual data flow programming and imperative programming. The novice user initially develops his data flow program through the familiar visual programming environment. This environment is effectively an intuitive user interface masking the underlying DesignScript language. The DesignScript language and its related user interface addresses three issues: the domain specific requirements of architectural and engineering computing, the scalability issues encountered when visual data flow programming is applied to complex design scenarios and the abstraction barriers encountered when users transition from data flow to imperative programming.

Concurrent programming is infamous for its difficulty. An important source of difficulty is non-determinism, stemming from unpredictable interleavings of concurrent activities. Futures and promises are widely-used abstractions that help designing deterministic concurrent programs, although this property cannot be guaranteed statically in mainstream programming languages. Deterministic-by-construction concurrent programming models avoid this issue, but they typically restrict expressiveness in important ways. This paper introduces a concurrent programming model, Reactive Async, which decouples concurrent computations using so-called cells, shared locations which generalize futures as well as recent deterministic abstractions such as LVars. Compared to previously proposed programming models Reactive Async provides (a) a fallback mechanism for the case where no computation ever computes the value of a given cell, and (b) explicit and optimized handling of cyclic dependencies. We present a complete implementation of the Reactive Async programming model as a library in Scala. Finally, the paper reports on a case study applying Reactive Async to static analyses of JVM bytecode based on the Opal framework. QC 20161219