Racket's Rhombus 1.0 turns a decades-old macro thesis into a cleaner language bet

Why it matters

Rhombus is a reminder that not every important developer-platform bet comes from a startup. Racket's core contributors are testing whether language extensibility can matter again if the syntax barrier is removed.

Matthew Flatt, Robert Bruce Findler, Sam Tobin-Hochstadt and the Racket community released Rhombus 1.0 on June 22, turning a long-running programming-languages research project into a production-intended language built for developers who want Racket's extensibility without writing Lisp-style parenthesized code.

That makes Rhombus an unusual release in the current developer-tool market. There is no startup founder, valuation, seed deck or enterprise sales motion behind it. The builders are programming-language researchers and open-source contributors. The Racket blog lists 20 major contributors, including Mashfi Ishtiaque Ahmad, Taylor Allred, Nia Angle, Wing Hei Chan, Stephen De Gabrielle, Findler, Jacqueline Firth, Flatt, Oliver Flatt, Kiran Gopinathan, Ben Greenman, Siddhartha Kasivajhula, Alex Knauth, Jay McCarthy, Lucas Myers, Alec Mills, Sam Phillips, Sorawee Porncharoenwase, Jens Axel Sogaard and Tobin-Hochstadt.

The senior names matter because Rhombus is less a new-language flyer than a continuation of the Racket thesis by its own core practitioners. Flatt is a professor at the University of Utah's Kahlert School of Computing whose research interest is programming-language implementation. Findler is a Northwestern computer science professor whose work spans programming languages, development environments and software contracts. Tobin-Hochstadt is an Indiana University associate professor and director of undergraduate studies for computer science, with research interests listed in programming languages, systems and high-performance computing.

What Rhombus changes

Rhombus is a general-purpose programming language built on Racket. The Racket announcement describes it as functional, extensible and dynamic, with interactive flexibility plus static and abstraction-enforcing constructs meant to scale from small scripts to larger systems. The practical hook is syntax: Rhombus keeps Racket's language-extension model, but presents it through more conventional notation.

That sounds cosmetic until you look at what Racket is actually known for. Racket has long positioned itself as a language for making languages. Its own homepage emphasizes a mature cross-platform runtime, a package system, a GUI framework, standalone binaries, a foreign-function interface, a web server, database support, math and statistics libraries, DrRacket and thousands of additional packages. Its differentiator, though, is macro-powered language construction: programmers can define domain-specific languages as libraries rather than waiting for the core language to absorb every abstraction.

Rhombus is the answer to the adoption tax in that model. The goals page, copied into the release announcement, says modern languages already agree on basics such as lexically scoped variables, closures, objects, pattern matching and type parametricity. The gap Rhombus targets is the layer above that: language support tailored to specific domains, without forcing developers to buy into Lisp's parenthesis-heavy notation.

The project says Rhombus is not simply Racket with different punctuation. The release FAQ says #lang rhombus improves on #lang racket with better predefined data structures, especially lists, a new class system, pervasive pattern matching, extensible static information, hierarchical namespaces and more. It also says Rhombus is built on Racket, uses DrRacket and the raco command-line suite, and relates to Racket roughly the way Elixir relates to Erlang or Kotlin relates to Java.

A research project reaches the 1.0 line

The 1.0 release follows public academic work that framed Rhombus as a way to get macro power without Lisp notation. A 2023 OOPSLA paper, "Rhombus: A New Spin on Macros without All the Parentheses", was authored by Flatt, Taylor Allred, Nia Angle, Stephen De Gabrielle, Findler, Jack Firth, Kiran Gopinathan, Ben Greenman, Siddhartha Kasivajhula, Alex Knauth, Jay McCarthy, Sam Phillips, Sorawee Porncharoenwase, Jens Axel Sogaard and Tobin-Hochstadt.

The paper's core argument is the same bet now packaged for users: many languages have macro systems, but Racket's style of extensibility remains unusually expressive; Rhombus tries to expose that power through infix notation, context-specific sublanguages, pervasive pattern matching, repetition notation and support for carrying static information through expansion.

The 1.0 label also matters because Racket's announcement draws a line between research and usable software. The FAQ says Rhombus is rooted in academia, but is not a teaching language and is not just a research language; it is intended for production use. That is a deliberately stronger claim than a paper artifact or experimental dialect.

Still, Rhombus is not entering the world with the network effects of Python, TypeScript, Rust or Go. Its strongest distribution advantage is inherited: because Rhombus is an outgrowth of Racket, it can lean on Racket's toolchain, package infrastructure, DrRacket and long-running community rather than starting from zero.

The open-source economics are different

Rhombus also shows what a serious language launch looks like outside venture-backed developer tooling. The relevant institutional backstop is the Racket Programming Language Foundation, which says it is registered in Delaware and recognized as a U.S. 501(c)(3) public charity. The foundation says donations support community infrastructure, administration, educational outreach and events such as RacketCon and Racket School.

That means the incentives are not the usual ones for a devtools startup. Rhombus does not need to convert free users into paid seats or justify a valuation. It does need to attract contributors, documentation energy, package authors and enough production users to prove that its macro-first approach is more than an elegant academic stance.

The public repository gives a modest but real signal of interest. The racket/rhombus GitHub repository showed 543 stars, 72 forks and 2,472 commits when checked. The repository also tells users they can install a Rhombus distribution, install the rhombus package from DrRacket, or use raco pkg install rhombus from the command line.

The AI-era wrinkle

The most current part of the Rhombus pitch is not syntax. It is the claim that domain-specific languages may become more valuable, not less, as coding agents become part of programming work.

The release FAQ addresses the question directly: do new languages or DSLs still matter in an age of autonomous coding agents? The project's answer is cautious but pointed: early predictions that AI would end new languages have not panned out, and the FAQ argues that raising the level of discourse in programming can help both humans and coding agents.

That is the real bet under the release. Rhombus is not trying to win by being the next default systems language or the next web-app workhorse. Flatt, Findler, Tobin-Hochstadt and the broader contributor group are betting that software teams will still need better ways to create smaller languages inside larger systems, and that the syntax burden that kept many developers away from Racket-style extensibility can be reduced without giving up the underlying power.

For now, that is a product promise more than an adoption story. Rhombus 1.0 has documentation, downloads, examples and a serious lineage. What it does not yet have, at least in the public materials, is a list of production users, a governance breakdown specific to Rhombus, or a clear measure of real-world usage beyond repository activity and community participation. For an open-source language, those are the metrics that will matter after the release-week curiosity fades.