PAINT '25: Proceedings of the 4th ACM SIGPLAN International Workshop on Programming Abstractions and Interactive Notations, Tools, and Environments

Some things are better drawn than coded. However, it takes a lot of work to build a custom editing interface for each new notation, discouraging experimentation. There has to be a better way than Greenspunning poor approximations of Adobe Illustrator over and over again. "Self-raising diagrams" are a promising escape from this trap. Just as source code — a static artefact — "raises itself" into a dynamic running program, a vector graphics diagram — taken to generalise source code — can similarly be parsed, interpreted, and "animated" into a GUI. Notational engineers can then focus on the notations themselves, and their semantics, having left the implementation of drawing interfaces to the experts (implementors of standard vector graphics editors). The tasks of normalising, interpreting, and transforming vector notations contain many relevant problems for the notational engineer — far more relevant than coding line rubberbanding for the umpteenth time!

Block-based programming environments offer unique benefits for editing or integration of visual tools that could be useful across programming environments. However, most general-purpose programming environments are designed around textual representations of programs.

We explore the design of a block-based editor that integrates with existing textual environments, using the example of the Squeak/Smalltalk development environment. Through a user study, we show that users felt generally comfortable with our editor’s edit interactions when compared to textual editing. We discuss the design's difficulties and opportunities we observed during our user study and our own use to help propose block-based editor designs that integrate well with text-based environments.

Programmers ask complex questions in their search for solutions during software development. Along with traditional tools such as debuggers and profilers, state-of-the-art approaches like Babylonian Programming can help programmers answer those questions through interactive and visual feedback. Large language models (LLMs) and programming agents are part of programmers' toolboxes and are well-integrated into their development workflows. However, they are not yet helpful in considering questions involving run-time behavior.

In this paper, we first review the literature to identify concerns programmers face during development and highlight how humans usually address them. We then focus on questions about program behavior and propose integrating Babylonian-style programming techniques with LLMs to help answer related questions. Finally, we suggest four key properties that future LLM-based development tools should support: (1) LLM tool usage traceability for explainability, (2) resumability of development progress for handovers between human programmers and LLM-based programming agents, (3) context efficiency through selective data querying, and (4) multi-source synthesis for tool integration.

The PyTamaro approach to introductory programming equates composing a program to composing a graphic. Using the simple PyTamaro library for Python, beginner programmers compose function calls that produce primitive graphics, such as rectangle(20, 10, red), with function calls that combine graphics into composites, such as above(_, _). To ease students into programming, several school teachers have been using TamaroCards, a paper-based visual language that can express the subset of Python needed to compose simple graphics with PyTamaro. This paper introduces TIDE, the TamaroCards IDE, a web-based environment to complement the unplugged, paper-based TamaroCards. Given the constraints imposed by the Python programming language, the PyTamaro library, and the TamaroCards notation, the paper explores the design space for an interactive TamaroCards programming environment, describes a design that satisfies the pedagogical needs, and evaluates the implemented design using the Cognitive Dimensions framework.

The complexity of functional programming languages can pose a challenge for learners. However, the use of block-based languages in learning programming can lower the barriers to the learning process. While many block languages have been created, they often lack essential features of functional languages and do not include the type inference.

This gap serves as motivation to develop a block-based functional programming language that provides visual information about three programming language conventions. The development process begins with designing text syntax, transforming it into blocks, drafting typing rules for visual languages, and building a new functional block language called Macaca Nigra Language.

Case studies of Macaca Nigra Language demonstrate that it effectively provides visual clues through shapes and colors regarding the three conventions of programming languages.