The lost cause of the Lisp machines

Original Article ↗ Hacker News Discussion ↗

Business and Hardware Decisions

  • Several comments argue Symbolics misidentified its “special sauce” as custom CPUs instead of the Genera environment, delaying a serious move to commodity hardware.
  • Some think a full native port to 80386 PCs (or similar 32-bit CPUs) could have matched performance “well enough” at a fraction of the cost, unlike Open Genera’s Ivory-emulator approach.
  • Others are skeptical this would have saved them, noting Xerox/Interlisp and Venue tried similar ports and mostly ended up as legacy-support vendors while the entire expert systems/Lisp market dried up.
  • DEC Alpha’s “open” marketing and OpenVMS/OpenGenera naming triggered a side discussion: “open” then mostly meant standards, networking, and POSIX, not open source.

Romanticism vs Realism

  • The article’s impatience with “Lisp machine romantics” drew pushback: many value these systems as a “vision of a future that never happened,” similar to Amiga, 8‑bit, PDP‑10, and Smalltalk nostalgia.
  • Defenders say this romanticism preserves history and highlights lost ideas: fully integrated, introspectable systems where “everything is an object” and the environment is deeply live and debuggable.
  • Critics counter that Lisp machines couldn’t reliably produce shippable, reproducible products; every machine became a bespoke lab, badly aligned with the emerging packaged-software economy.

Environment, Tooling, and Live Systems

  • Multiple anecdotes praise the “Lisp all the way down” experience: live object graphs behind the UI, crash recovery by editing the running image, time-travel debugging, and sophisticated integrated tools (search/replace across systems, source compare, etc.).
  • Some see modern Common Lisp environments, Emacs, Clojure, Racket, and projects like “freestanding Lisps on Linux syscalls” as partial spiritual successors, but acknowledge nothing matches Genera’s depth of integration.

Interoperability and Shipping

  • Historically, Lisp ecosystems often assumed interactive hacking over productization, which some participants directly link to Lisp machines’ commercial failure.
  • There’s a parallel drawn to today’s scripting languages and fragile packaging (especially Python), contrasted with FP and Lisp systems that do emit robust executables or jars.
  • Others note Lisp’s interoperability has improved substantially thanks to stable C ABIs and numerous Lisps that target mainstream runtimes (JVM, .NET, BEAM, JS, etc.).

Specialized Hardware and AI Parallels

  • Several comments generalize Lisp machines to a broader pattern: specialized hardware cycles (word processors, transputers, graphics, AI accelerators) repeatedly get overtaken by general-purpose systems.
  • On AI, commenters agree the technology will persist but expect an eventual “shakeout” where many GPU-heavy AI companies fail, leaving surplus specialized hardware—echoing the Lisp machine era.