Morten opens DYNA26 with an overview of the agenda and an updated road map.

At DYNA Fall 2025, Alex Holztapple from METSIM International presented the new version of METSIM®, which is based on Dyalog APL. The tools that were developed to support this project are now documented and available free of charge to anyone who would like to migrate APL code to Dyalog APL from other APL implementations. Morten shows you where to get these tools and how to use them.

Dyalog provides a toolset that enables APL developers to work alongside other technologies. Rich walks through an application that combines several of these tools – Jarvis for HTTP serving, SQAPL for database integration, and isolates for concurrent computation – to build a video search and recommendation engine with a React frontend.

Mark and Kori share a data processing challenge from the jewellery industry – normalising unstructured product descriptions into clean data that can be reliably queried. They compare approaches using large language models (LLMs) and APL.

Over the last twenty years, Dyalog has implemented many useful extensions and generalisations of the APL primitive set. Asher presents some “highlights” of these enhancements, demonstrating ways in which you can make your code more concise and performant. He discusses leading axis theory, new primitives, and new ways to use old primitives.

Asher presents some of the enhancements that were in introduced in Dyalog v20.0:

• Array notation introduces a literal syntax for most arrays – including nested and high-rank arrays – and namespaces, enabling direct construction and manipulation of complex structures.
• Namespace handling is streamlined with ⎕VGET, ⎕VSET, and extensions to ⎕NS. These are particularly useful when working with namespaces produced by ⎕JSON from web-service responses, allowing default values to mitigate missing members.
• The new compositional operator ⍛ (behind) captures the common patterns (f Y) g Y and (f X) g Y, improving clarity and reducing boilerplate.
• Inline tracing enables primitive-by-primitive evaluation with access to arguments and results, supporting debugging, teaching, and performance analysis.

Web APIs are everywhere, but consuming them in APL typically means hand-drafting repetitive boilerplate. OpenAPI is a widely-adopted standard for describing REST APIs in a machine-readable format. Holden presents a new tool that uses OpenAPI specifications to automatically generate APL client code built on top of HttpCommand.

The exponential growth in the pervasiveness and power of AI led Brian to ask what would happen if he tried to use AI to generate a web front end for, or perhaps even build completely, a Jarvis-based web service. Could AI improve the speed of development, or relieve some of the tedium of building UI? Brian shares the results and lessons learned from his efforts.

Kyle shares what he learned when building a Minecraft-style voxel game using Dyalog. He explains how APL's array-oriented primitives and terse notation turned complex game engine problems into sharable, screen-sized, snippets of code. He also describes the experience and challenges involved in learning APL from a newcomer's perspective.

Are you passionate about growing the next generation of APLers? Do you have ideas for innovative projects that showcase the power of APL in science, engineering, or mathematics, but lack the funding to bring them to life?

Mark provides an update on The APL Trust, a non-profit charity dedicated to supporting projects that promote global awareness, and meaningful use, of the APL programming language. He explains how the APL Trust can help turn your ideas into impactful initiatives that advance education and technical excellence through APL.