HOL (theorem prover)

HOL (theorem prover)
Name	HOL
Developed by	Cambridge University, University of Cambridge Computer Laboratory, University of Cambridge, University of Cambridge Computer Laboratory alumni and collaborators
Initial release	1980s
Latest release	multiple variants maintained continuously
Programming language	Standard ML, OCaml, ML implementations
Operating system	Unix-like, Linux, macOS
License	various academic and open-source licenses

HOL (theorem prover)

HOL is a family of interactive proof assistants originating in the 1980s for machine-checked reasoning about higher-order logic. It has been used across academic and industrial projects for formal verification, theorem proving, and mechanized mathematics, influencing developments in proof engineering, software verification, and hardware certification.

History and development

The HOL family emerged from work at institutions such as University of Cambridge, University of Cambridge Computer Laboratory, Imperial College London, and collaborations with researchers at Stanford University, Carnegie Mellon University, and MIT. Early contributors included members of research groups associated with projects at Technische Universität München and University of Cambridge Computer Laboratory alumni who adapted ideas from systems like LCF (logic for computable functions), Edinburgh LCF, and efforts at Cambridge University to create an environment centered on a small trusted kernel and programmable automation. As interest spread, variants and forks appeared at places such as University of Cambridge, Middlesex University, University of Cambridge Computer Laboratory, University of Oxford, and companies collaborating with Intel Corporation and ARM Holdings, while workshops and conferences like CADE, IJCAR, and CPP (Conference on Certified Programs and Proofs) showcased developments. Funding and coordination involved grants from organizations such as Engineering and Physical Sciences Research Council and partnerships with laboratories including National Physical Laboratory, leading to an ecosystem of implementations and community-maintained libraries.

Architecture and implementations

The HOL family follows the LCF-style architecture with a small trusted core implementing inference rules and a programmable metalanguage, typically Standard ML or OCaml. Implementations include prominent systems developed at University of Cambridge, at SRI International and affiliated teams, and at groups connected to University of Cambridge Computer Laboratory alumni; notable variants are used in projects at INRIA, University of Cambridge, and industry partners like Intel Corporation. The architecture separates a kernel for proof object construction from higher-level tactics and decision procedures, enabling extensions at places including Microsoft Research, Bell Labs alumni projects, and university research groups. Users interface with the system via interactive shells, batch scripts, or graphical front-ends produced by developers at University of Cambridge and collaborators at University of Oxford, while platform support spans systems used at Los Alamos National Laboratory and academic departments across United Kingdom and United States institutions.

Logic and semantics

HOL systems formalize classical higher-order logic with typed lambda calculus foundations inspired by work at Princeton University and University of Edinburgh groups. The logic supports simple type theory, polymorphism, and definitional extension principles developed in research programs at Cambridge University, Stanford University, and Carnegie Mellon University. Semantic models and soundness proofs draw on results from logicians at University of Cambridge and University of Oxford, and on theories connected to the work of scholars at University of Michigan and Harvard University. This logical framework has enabled formal treatment of arithmetic, set-theoretic encodings, and domain theory used in projects at INRIA, SRI International, and University of Cambridge Computer Laboratory. The design emphasizes a minimal trusted kernel to reduce the TCB, an approach advocated by researchers associated with Edinburgh LCF and later propagated at venues like CADE and IJCAR.

Formalization libraries and theories

Extensive libraries accompany HOL variants, covering algebra, analysis, topology, measure theory, and discrete mathematics—work developed by contributors at University of Cambridge, INRIA, University of Oxford, University of Cambridge Computer Laboratory alumni, and collaborative efforts with groups at MIT and Carnegie Mellon University. Formalizations include mechanized proofs of classical theorems influenced by teams at Princeton University and Harvard University, and certified libraries for number theory, probability, and real analysis used in projects at National Physical Laboratory and Los Alamos National Laboratory. Large-scale theory repositories hosted by universities such as University of Cambridge and institutes like INRIA provide reusable components for verification of systems designed by industry partners including ARM Holdings and Intel Corporation.

Tools, interfaces, and automation

The HOL ecosystem features tactic languages, proof automation, counterexample generators, and integrations with model checkers and SMT solvers developed in collaboration with researchers at Microsoft Research, SRI International, and INRIA. Interactive environments and IDEs emerged from projects at University of Cambridge and University of Oxford, while bridging tools connect HOL to external systems like Isabelle and proof exchange standards discussed at conferences such as CADE and CPP (Conference on Certified Programs and Proofs). Automation components—decision procedures, rewriting engines, and reflective tactics—were advanced by groups at Carnegie Mellon University, Stanford University, and University of Cambridge Computer Laboratory alumni, enabling large proofs in hardware verification and certified compilation efforts supported by organizations such as DARPA and national research councils.

Applications and case studies

HOL systems have been applied to verification of processors, microkernels, compilers, and cryptographic protocols in projects involving ARM Holdings, Intel Corporation, GCHQ, and academic partners at University of Cambridge and University of Oxford. Case studies include verified hardware designs used in collaborations with National Physical Laboratory and proofs of correctness of system software produced by teams at SRI International and Carnegie Mellon University. Formal proofs of mathematical theorems and certified numerical algorithms have been published by researchers affiliated with INRIA, Princeton University, and Harvard University, while industrial adoption in safety-critical domains has been promoted through partnerships with standards bodies and laboratories such as Los Alamos National Laboratory and National Physical Laboratory.

Category:Theorem provers