Curated Picks

A podcast series featuring engineering leaders from CyCognito, Metabase, Red Planet Labs, JustOn, Mobot, and HolidayPirates offers a cross-section of long-term Clojure usage in production. At CyCognito, Yehonathan Sharvit describes a system processing millions of events per second, but notes that most engineers still favor more mainstream languages for career reasons, which has driven a partial migration toward TypeScript. In contrast, Nathan Marz argues that Rama's roughly 200,000 lines of Clojure, representing a decade of R&D, would have been impractical to build in other languages due to its reliance on macros and enforced immutability.

Cam Saul from Metabase highlights a different dynamic, pointing to successful contributor-to-employee conversion via their open-source project and the ability to ship enterprise-scale features with teams of four as a sustained advantage. Across the interviews, a consistent theme emerges: moves away from Clojure are driven primarily by hiring and market perception rather than technical limits, while teams that stick with it tend to emphasize small team productivity as the key factor.

A February 2025 Hacker News discussion of a "Why Clojure?" essay generated over 200 comments debating the language's production viability. Practitioners reported 10+ year codebases with minimal dependency churn and cited REPL-driven development as enabling faster iteration than compile-test-debug cycles. Critics raised hiring difficulties, dynamic typing friction in large codebases, and community fragmentation around tooling (lein vs deps.edn vs boot). Multiple commenters noted teams initially adopting Clojure for greenfield projects, then migrating to Go or Kotlin after key developers departed, attributing abandonment to perceived elitism and tooling complexity rather than technical deficiencies.

A 2020 ClojureVerse thread debated whether ClojureScript/Node.js should become Clojure's primary runtime given JavaScript's dominance. Core maintainers (dnolen, thheller, seancorfield) rejected the premise, citing JVM superiority for server workloads, multi-threading, and production stability. The creator of Shadow-cljs noted most users were "surprised" the tool required a JVM, suggesting the platform wasn't the adoption barrier. Five years later, Clojure remains JVM-centric with minimal ClojureScript server adoption, validating maintainers' assessment that JavaScript developers weren't blocked by tooling but by the language's steep learning curve and functional paradigm shift.

In March 2020, a ClojureScript discussion revisited whether Phoenix LiveView's server-driven UI model represented a serious threat to the ecosystem. Thomas Heller, the maintainer of shadow-cljs, framed LiveView as one option on a broader trade-off spectrum rather than a fundamental shift, while also pointing out how heavily ClojureScript had converged on React as a single dominant solution. Much of the debate focused on practical concerns: diffing performance, websocket behavior behind corporate proxies, and the balance between server-managed state and rich client-side interaction. By May 2020, Tatu Tarvainen released Ripley, a Clojure implementation of the LiveView approach.

In January 2018, a ClojureVerse discussion surfaced Eric Normand's tweet questioning whether Clojure offered compelling web development advantages. Community members attributed adoption barriers to tooling complexity (setup friction, fragmented documentation, learning curves requiring JVM/JS expertise), paradigm unfamiliarity (functional programming and structural editing presented friction for developers from imperative/OO backgrounds), and communication failures (inability to articulate benefits, lack of "batteries-included" frameworks comparable to Rails).

In October 2017, a thread on ClojureVerse questioned whether Clojure's appeal skewing toward "grumpy old programmers" reflected an outreach issue, noting the low number of beginners showing up at conferences after a decade of availability. Much of the discussion centered on tooling friction, especially the complexity of getting a usable IDE setup, typically involving IntelliJ with Cursive or a nontrivial Emacs configuration. One newcomer described spending days just getting builds working, in contrast to other languages that were usable within minutes.

The thread also touched on whether ClojureScript was mature enough to warrant more active outreach to JavaScript developers. At the time, platforms like Egghead.io offered extensive React and Elm material, but no ClojureScript courses, reinforcing its relative invisibility.