FARM '25: Proceedings of the 13th ACM SIGPLAN International Workshop on Functional Art, Music, Modelling, and Design

TidalCycles is a library for live-coded music performance based in Haskell, and Strudel is an adaptation of TidalCycles in JavaScript. Both have seen wide use, but their implementation within existing environments for evaluating code limits the ways in which a live code editor can integrate with the compilation process. This demo presents Weft, a web-based Haskell-like interpreter which allows performers to use TidalCycles syntax and type definitions to control Strudel's library of JavaScript pattern functions. By implementing the parser, type checker, and runtime in a JavaScript module, Weft enables interactive use in the browser, while also adding type inference and error reporting rarely seen in web-based live coding environments.

We present Miniblazon, a simplified version of blazon, the language historically used to describe coats of arms. We introduce a type system that prevents several forms of invalid blazons as well as violations of stylistic conventions. We implement the system as a command-line application and use it to analyze a corpus of historical blazons, identifying several as convention-breaking.

Rocq N’Roll is a software instrument and tool with type proof -> music. It parses Rocq proof tactics into Musical Instrument Digital Interface (MIDI) notes to generate musical sound based on a proof state. ’Playing’ or ’performing’ a proof with Rocq N’Roll means stepping through it interactively to generate pitches that correspond to the step in the proof, with more complex chords and melodies coming from proof steps with multiple suboperations (i.e. automated tactics). Rocq N’ Roll is primarily an artistic tool, but it may also have interesting use cases in helping Rocq users conceptualize proof deautomation. And (in our humble opinion) it sounds pretty sweet.

This paper introduces an algorithmic performance approach for transforming real-time environmental data into collaborative, generative sound practices. Drawing from atmospheric and geospatial parameters — including temperature, humidity, and pollution indices — the system generates both electronic soundscapes and adaptive notations for acoustic performers. Through live navigation across global environmental conditions, the framework explores dynamic, location-specific harmonic structures, integrating cultural and ecological dimensions into the compositional process. By combining real-time data acquisition, functional programming techniques, and algorithmic composition, the system establishes a dialogic relationship between computational processes and human performers. Departing from traditional sonification models based on symbolic or analytical mappings, this approach frames live performance as a hybrid artistic space where environmental data actively shape musical expression and audience experience. The work aims to position environmental sonification within the broader context of generative art and real-time algorithmic performance, fostering new modes of artistic and ecological engagement.

I present a translation of Girard's paradox into music using Soundproof, an in-progress system for translating proof terms of dependently typed lambda calculus into electronic music according to their tree structure. I explore methods of representing tree structures sonically, and points of choice in the translation and presentation of the result.

The aim of this work is to offer a reflective auditory environment that prompts contemplation about the direction of technology and music. The multiple drones’ consistent presence symbolizes the steady advancement of technology, while the ambient layers depict the evolving nature of music. The drones and tones combine with a spoken word description of ‘cyberspace’ first documented by William Gibson.

In this article, we provide a generalization of turtle graphics that can be specialized to create various domain-specific languages. In particular, we provide two specializations of our language: the first one produces Bézier paths and the second one combines and manipulates these paths for vector graphics. Additionally, we introduce a syntax that adds absolute positioning to the relative positioning of turtle graphics.

We demonstrate how a systematic way to extend the turtle's state allows generic equational reasoning on programming languages similar to turtle graphics as well as a rich way to specify languages for media arts.

This demo presents an algorithmic performance system that transforms real-time environmental data into generative soundscapes and live performance materials. Built at the intersection of algorithmic composition, live coding, and environmental sonification, Sonic Earth demonstrates how planetary conditions --- including temperature, humidity, and pollution indices --- can shape collaborative musical practice in real time. This demo builds on a system recently selected for the demo showcase at the International Computer Music Conference (ICMC), expanding its presentation towards FARM's focus on functional, generative, and performative dimensions. The system integrates electronic sound synthesis, adaptive graphic notation, and human performers, establishing a dialogic performance space where environmental parameters directly influence musical structure, sonic texture, and interpretative gestures.

Embedded music programming languages confer a number of benefits to composers by bringing music objects into the world of modules, bindings, and code reuse. However, many composers write music in graphical score editors. A natural desire, then, would be to embed music languages into these editors. A proper embedding would provide a means to export the score-with-embedded-code from the score editor to a program in the music language's host language. This program ought to be loadable into a larger program context, leveraging the modular features of the host language to share bindings with other host language modules.

That is the subject of this demo. The demo will use the MuseScore score editor and the Tonart music language embedded in the Racket language, but the same principles can be applied in other score editors, or with other choices of music and host language.