{"id":570277,"date":"2026-05-07T02:10:13","date_gmt":"2026-05-07T02:10:13","guid":{"rendered":"https:\/\/www.newsbeep.com\/uk\/570277\/"},"modified":"2026-05-07T02:10:13","modified_gmt":"2026-05-07T02:10:13","slug":"ruby-inventor-matz-working-on-native-compiler-with-ai-help","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/uk\/570277\/","title":{"rendered":"Ruby inventor Matz working on native compiler with AI help"},"content":{"rendered":"

DevOps<\/p>\n

Matz gets together with Anthropic’s Claude to create an experimental ahead-of-time compiler for Ruby \u2013 though with many limitations<\/p>\n

Yukihiro Matsumoto – better known as Matz – is building Spinel, a native compiler for Ruby, with help from Anthropic’s Claude Code.<\/p>\n

Spinel, which is on GitHub<\/a>\u00a0under the MIT license, works by parsing Ruby code into AST (abstract syntax tree)
\nfiles, then converting it to C code for compilation by a standard C
\ncompiler. In Matz\u2019s tests, Spinel-compiled code runs approximately 11.6
\ntimes faster than MiniRuby –\u00a0a stripped-down Ruby build – using the in-development Roby 4.1.0.\u00a0<\/p>\n

The output is C code, compilable to a native executable via gcc (GNU Compiler Collection) on Linux or Windows (with MinGW – Minimalist GNU for Windows), or via LLVM’s Clang on Linux or macOS. BSD will “probably work”, according to the readme, but is not tested.<\/p>\n

Ruby is an interpreted language, meaning it depends on
\na runtime engine to parse and run the code. In order to improve performance,
\nRuby can use just-in-time (JIT) compilers including MJIT (method-based JIT), YJIT
\nand ZJIT, with the latter two developed by Shopify, a prominent Ruby user.
\nSpinel is different in that it generates standalone native code executables,
\nwhich can be deployed without any additional runtime.<\/p>\n

The downside of Spinel is that it supports only a subset of
\nRuby. Unsupported features include eval statements, which evaluate and execute
\nRuby code at runtime, threads, text encoding other than UTF-8, metaprogramming
\nsuch as defining a method at runtime, and deeply nested lambda functions.<\/p>\n

Ruby variables are not typed, though objects are strongly
\ntyped. Spinel performs type inference to enable C code generation, since C is a
\nstrongly typed language. The code makes use of an existing and mature Ruby
\nparser called Prism. There are many optimizations, such as method inlining and
\ndead code elimination, and the generated C compiles cleanly at the default
\nwarning level. Spinel includes a garbage
\ncollector, to reclaim memory automatically, and supports FFI (foreign function
\ninterface) for integrating with native code libraries such as libc or SQLite.<\/p>\n

Spinel is experimental and will not work with most existing
\nRuby code, including the web application framework Ruby on Rails. But it is
\npossible for Ruby developers to write code with Spinel in mind, such as
\nfor helper functions that can then be called from other Ruby code, as a means
\nof optimization.<\/p>\n

Matz presented Spinel at RubyKaigi 2026, a conference in
\nHakodate, Japan last month. According to attendees such as this one<\/a>\u00a0(original text in Japanese) Matz said the idea for Spinel was conceived three years
\nago, but has now been implemented in a few weeks using AI.\u00a0<\/p>\n

Most of the code in
\nthe Spinel repository is headed with a comment including “co-authored by:
\nClaude Open 4.7 (1M context).”\u00a0<\/p>\n

Further, the project has already been rebuilt three
\ntimes over, in a series of experiments.<\/p>\n

Matz, perhaps, is an ideal user of AI-generated code. He
\nunderstands the code, and can benefit from the increased speed of development
\nwithout losing grip on what the code does, and is able to refine it with his existing
\nskills as well as with further AI prompting, and the code is covered by hundreds
\nof tests and benchmarks. \u00ae<\/p>\n","protected":false},"excerpt":{"rendered":"DevOps Matz gets together with Anthropic’s Claude to create an experimental ahead-of-time compiler for Ruby \u2013 though with…\n","protected":false},"author":2,"featured_media":570278,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[554,733,4308,86,56,54,55],"class_list":["post-570277","post","type-post","status-publish","format-standard","has-post-thumbnail","category-artificial-intelligence","tag-ai","tag-artificial-intelligence","tag-artificialintelligence","tag-technology","tag-uk","tag-united-kingdom","tag-unitedkingdom"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/posts\/570277","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/comments?post=570277"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/posts\/570277\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/media\/570278"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/media?parent=570277"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/categories?post=570277"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/uk\/wp-json\/wp\/v2\/tags?post=570277"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}