PLSS 2026

Workshop on Programming Language Standardization and Specification

This workshop aims to foster cross-pollination between researchers and industry professionals with experience in programming language specification and standardization. It provides a forum where participants can share insights, case studies, and best practices, and collaboratively explore solutions to current challenges.

The goal of the workshop is to improve the collective understanding of how programming languages are specified, standardized, and evolved in practice.

The workshop examines specifications as the foundation of standards documents and language ecosystems, covering topics such as mechanized specifications, socio-technical aspects of language evolution, decision-making processes in standardization, intellectual property considerations, and implications for language adoption.

Whether addressing historical perspectives, emerging trends, cross-disciplinary approaches, or novel methodologies, the workshop seeks to provide a platform for exploring diverse experiences.

Call for Presentations

The Workshop on Programming Language Standardization and Specification invites submissions for presentations that explore the specification and standardization of both widely adopted and emerging programming languages.

We welcome contributions that address a wide range of topics related to programming language specification and standardization, including but not limited to:

• programming language specification frameworks
• socio-technical aspects of programming language evolution, specification, and standardization
• approaches to expressing and documenting rationale during programming language evolution
• leveraging programming language standards to ensure compatibility of multiple implementations
• industrial implementation of programming language standards
• implications of standardization for language adoption
• empirical studies on programming language specifications
• interactivity, accessibility, and customizability of programming language specifications
• formal methods in programming language specifications
• mechanised and executable specifications
• intellectual property rights issues and legal aspects in programming language standardization
• “standardese” and linguistic aspects of programming language specification texts
• integrating programming language standardization in CS curricula

Submission Guidelines

Submission Link: https://plss2026.hotcrp.com

Prospective presenters should submit an abstract (2-5 paragraphs) outlining their proposed presentation. The abstract should clearly articulate the topic, key contributions, and relevance to the workshop themes. Submissions may include technical discussions, position statements, or experience reports. Submissions on work-in-progress are specifically encouraged. Previously presented or published work that is relevant to the workshop themes is also welcome.

Contributions from industry professionals, representatives of standardization bodies, and programming language implementors are especially welcome.

Submissions will undergo a lightweight review process.

This keynote will be given by Dr. Gilad Bracha, Technical Fellow, F5, and the recipient of the 2017 Senior Dahl–Nygaard Prize.

Specifying Languages and VMs: Looking Back and Looking Forward

Looking back, I’ve had a hand in several programming language specifications: notably, the Java Language and VM Specifications, as well as Dart, Newspeak, Vibe and others. These were primarily natural language specifications. We experimented with more formal approaches, notably encoding the JVM byte code verification rules in Prolog, but formal specification remained on the margins. Why did we eschew formal specifications? Academics worked on formal specs, with limited success. They usually glossed over crucial complications: class loaders, reflection, serialization, access control. And formal specs were not very accessible to implementors of compilers and VMs. Last but not least, the ambiguity of natural language gives wiggle room when interpretations need to be quietly changed. Can we do better? Looking forward, Artificial Intelligence opens up new possibilities of language specification. AI can be used to disambiguate natural language; it can produce formalizations, or copiuous examples that can accompany natural language specs and disambiguate them. In principle, an AI model can itself be the spec, and act as an oracle that implements it, answers questions about it, teaches etc. Ultimately, the AI writes a spec for (another?) AI to consume, and we can all go home.