Protocols are systems

Protocols are systems

A protocol is a relation over roles, so a protocol is where a system's systemhood lives.

R is defined over T
The arrow found, by position, in eight formal definitions of system.

Klir defines a system as a set of things \(T\) and a relation \(R \subseteq T \times T\). The relation cannot be stated without the things, and that dependence sits in the same position in all eight definitions encoded here. Whether it means the same thing in each is open. A protocol is rules, and rules are relations, so a protocol is the part of a system where systemhood resides. This project states that claim precisely, proves what can be proved in Lean, and keeps the rest as open problems.

30 statements · 16 formalized in Lean · 10 open problems

Contents

Contributing

Discussion takes place in SIGFPT, the formal protocol theory group of the Protocol Institute. Issues and pull requests are welcome on GitHub. The open problems in the blueprint are the natural places to start.

Building

lake exe cache get && lake build

The Lean files depend on systems-science-foundations, pinned in lakefile.lean. The blueprint is built with leanblueprint.

Changelog

How to cite

Thornton, S. (2026). Protocols are systems. Halcyonic Systems. halcyonic.systems/protocols-are-systems