kjnilsson/fez

Fez is a compiler and tooling suite that translates F\# code into Erlang Core and Beam files. It provides a command-line interface for building, testing, and releasing the compiler, alongside a standard library implementation for F\# modules and Erlang-style concurrency primitives. The system supports a wide range of F\# language features, including async workflows, pattern matching, and data types, enabling developers to write F\# code that compiles to efficient, distributed Erlang applications.

Added comprehensive implementations for F\# standard library modules, including \SeqModule\ (with functions like \scan\, \reduce\, \unfold\, \windowed\, \zip\, \groupBy\, \distinct\, \maxBy\, \minBy\, \sumBy\, \where\), \ListModule\ (with \compareWith\, \countBy\, \distinctBy\), \MapModule\, \SetModule\, and \OptionModule\. Introduced support for \FSharpAsync\ (including \Start\, \RunSynchronously\, \Parallel\), \Lazy\ evaluation, \Result\ types, and \Ref\ (mutable references). Added core operators such as \op\_Modulus\, \op\_GreaterThanOrEqual\, \op\_LessThanOrEqual\, and \byte\ conversion. Implemented \System.Char\ utility functions and basic object/inheritance dispatch mechanisms in \Fez.Core\.

src · high confidence

The Fez tooling now provides a command-line interface to compile F\# code into Erlang Core and Beam files. Users can run 'fez init' to scaffold a .NET project, 'fez compile' to generate .core files from F\# sources, and 'fez erl' to launch an Erlang shell with the correct search paths. The compiler leverages the F\# Compiler Service to parse and check F\# code, handling specific types like Option, Result, and List, and outputs Erlang modules with corresponding header information.

src/fez · high confidence

A new \Fez.Core\ module is added, providing core data types for Erlang-style distributed computing, including \Pid\, \Port\, \Node\, and a \Term\ union type that supports atoms, integers, floats, lists, bit strings, strings, maps, and PIDs. The module introduces basic process management functions such as \spawn\ and \self\, and a \send\ function that returns \unit\ rather than the message. It also adds support for reference cells (\Ref\<'a\>\) and lazy evaluation (\System.Lazy\<'a\>\) via extension members, alongside a \Binary\ type and helper module for byte array manipulation. Additionally, an \ErlangTerm\ attribute is introduced to control tuple generation and case strategies for F\#-to-Erlang term mapping.

src/fez.core · medium confidence

Adds a Makefile and platform-specific shell/Windows scripts (build.sh, build.cmd, run-tests.sh, run-tests.cmd) to automate the build, test, and release processes for the fez compiler. The Makefile defines targets for restoring .NET dependencies, compiling Erlang code, running EUnit tests, and creating self-contained releases for Windows, macOS, and Linux. Additionally, a ce.escript utility is added to format or display the core Erlang AST, and a thoughts.md document outlines the implementation approach for features like records, mutation, and discriminated unions.

(repo-wide) · high confidence

Added new test files (aether\_test.fs, basics.fs, demo.fs, demo2.fs, fun.fs, soduku.fs, wip.fs) and corresponding Erlang test modules (basics\_tests.erl) to verify the compiler's handling of records, pattern matching, tuples, lists, strings, discriminated unions, async messaging, and other core F\# constructs.

test · high confidence

The project files for fez, fez.core, and test have been updated to target .NET Core 2.1 and .NET Standard 2.0, replacing the previous .NET Core 2.0 and 2.1 targets. This ensures compatibility with the latest .NET Core runtime and standard libraries.

(dependencies) · high confidence

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

• CAI 45 → 48 (+2.8)
• Rubric changed (rubric-2026.08.19 → rubric-2026.09.15) — scores are not directly comparable.

• Code Health 85 → 85 (-0.1)
• Architecture 100 → 100 (+0.0)
• Maturity 37 → 49 (+12.3)
• Readiness 24 → 26 (+1.4)
• Security 100 → 87 (-13.3)

• Dependency hygiene not measured — dependency manifest found but not parsed for hygiene
• No exposed public API
• Test reliability not included
• TooManyMethods: Microsoft.FSharp.Collections.SeqModule (src/Microsoft.FSharp.Collections.SeqModule.erl)
• complexity unreadable for .erl, .fs — churn × complexity hotspots could not be measured
• dormant codebase — no living knowledge left to concentrate

• Documentation: no usage examples (README.md)
• Duplicated block (12 lines × 2) (src/Microsoft.FSharp.Collections.SeqModule.erl)
• End-of-life runtime: .NET netcoreapp2.1
• Microsoft.FSharp.Collections.SeqModule.next (cognitive 36) (src/Microsoft.FSharp.Collections.SeqModule.erl)
• Microsoft.FSharp.Collections.SeqModule.next (cyclomatic 95) (src/Microsoft.FSharp.Collections.SeqModule.erl)
• No ADRs found

• Unchanged — 0 containers · 1 contexts · 0 edges

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

kjnilsson/fez was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point a surveyor at a repository you know and see whether you agree with it.