This page is intended as a catalog of web-accessible information about Lisp Machines. Please submit your comments and links. Before you submit a link, do check the meta-pointers. This page is the product of your suggestions for incremental improvement; I no longer have the time to keep it up-to-date all by myself. I especially want to acknowledge Daniel Weinreb and Brian Mastenbrook for their comments.
Brad Parker wrote an emulator for the CADR, which is the second
model of MIT Lisp Machine, and the last that was designed fully at
MIT. It currently runs on GNU/Linux x86. Go check it
These are some websites that have a lot of Lisp Machine related content.
A few things I know about Lisp Machines.
This site has a large collection of links, covering virtually everything one could find by performing a Google search.
Lisp Machine supplies and information
Peter Paine, located in Kent, UK, actually sells used Lisp machines.
Symbolics Lisp Museum
The URL changed, so the link on Faré's site is borken.
A brief history of Lisp Machines
Includes an overview of most Lisp Machines ever built, including Interlisp machines, although it does not have the Japanese models.
A bibliography by Ralph Möller
In the beginning there was the AI lab, which mostly ran general purpose PDP-10 minicomputers with the Maclisp implementation of Lisp. Their story has been told elsewhere; here, I will just say that they developed the original CONS Lisp Machine as well as the CADR. The CADR technology was subsequently licensed to Symbolics and to Lisp Machine, Inc. (LMI). The period during which the Lisp Machine was a publicly funded AI lab project is particularly interesting, because there are detailed documents available that cover not only user-level documentation, but also design considerations.
AI WP-88; Tom Knight: CONS. November 1974.
AIM-444; Alan Bawden, Richard Greenblatt, Jack Holloway,
Thomas Knight, David Moon and Daniel Weinreb: LISP Machine
Progress Report, August 1977
"This informal paper introduces the LISP Machine, describes the goals and current status of the project, and explicates some of the key ideas. It covers the LISP machine implementation, LISP as a system language, input/output, representation of data, representation of programs, control structures, storage organization, garbage collection, the editor, and the current status of the work."
AIM-528; F. Knight, Jr., David A. Moon, Jack Holloway and Guy
L. Steele, Jr.: CADR, May 1979
"The CADR machine, a revised version of the CONS machine, is a general-purpose, 32-bit microprogrammable processor which is the basis of the Lisp-machine system, a new computer system being developed by the Laboratory as a high-performance, economical implementation of Lisp. This paper describes the CADR processor and some of the associated hardware and low- level software."
AIM-602; Dan Weinreb and David Moon: Flavors: Message
Passing in the Lisp Machine, November 1980
"The object oriented programming style used in the Smalltalk and Actor languages is available in Lisp Machine Lisp, and used by the Lisp Machine software system. It is used to perform generic operations on objects. Part of its implementation is simply a convention in procedure calling style; part is a powerful language feature, called Flavors, for defining abstract objects. This chapter attempts to explain what programming with objects and with message passing means, the various means of implementing these in Lisp Machine Lisp, and when you should use them. It assumes no prior knowledge of any other languages."
AIM-628; A. Moon: Chaosnet. June 1981
"Chaosnet is a local network, that is, a system for communication among a group of computers located within about 1000 meters of each other. Originally developed by the Artificial Intelligence Laboratory as the internal communications medium of the Lisp Machine system, it has since come to be used to link a variey of machines around MIT and elsewhere."
Tom Knight. Implementation of a List Processing Machine.
David Moon, Richard Stallman and Daniel Weinreb: Lisp
Machine Manual. Sixth Ed., System v. 99, June 1984.
This particular version of the Chine Nual is actually a fairly late version. I have not seen early Chine Nuals on the net. This edition includes the following historically interesting note by RMS:
"The Lisp Machine is the product of the efforts of many people too numerous to list here and of the former unique unbureaucratic, free-wheeling and cooperative environment of the M.I.T. Artificial Intelligence Laboratory. I believe that the commercialization of computer software has harmed the spirit which enabled such systems to be developed. Now I am attempting to build a software-sharing movement to revive that spirit from near oblivion."
More random bits and pieces about the CADR, including a
ZMail manual by RMS:
This is a set of pictures of an original CADR that is now in the MIT museum.
Lisp Machine Wikipedia entry
"Lisp machines were general purpose computers designed (often with hardware support) to efficiently run Lisp as their main language. In a sense, they were the first commercial single-user workstations."
LMI maintained very close relations with the AI lab; it is often hard to say whether a document comes out of LMI or MIT, and some LMI stuff might be found among the CADR documentation. After selling CADRs for a while, LMI developed the Lambda. Just before they folded, they built the K-machine, which was quite radically different from the original Knight Machine in that it was based on the "RISC" philosophy of microprocessor design.
Lisp Machine Inc. K-machine: The Deffenbaugh, Marshall,
Powell, Willison architecture as remembered by Joe
"The LMI K-machine was the last processor designed and built by Lisp Machine, Inc. Unlike other Lisp Machines, the K-machine is not descended from Tom Knight's original CONS architecture; the K-machine is an original design. This paper provides an overview of the more interesting elements of the K-machine architecture."
Of the two companies that the Lispm technology was licensed to, Symbolics was the most succesful. The fact that some of the following documents are patents is quite typical of Symbolics. Many interesting innovations came out of Symbolics, most notably in the areas of GUIs and graphics.
David A. Moon: Architecture of the Symbolics 3600. Proceedings of the 12th annual international symposium on Computer architecture, p. 76-83, 1985. Also published in: ACM SIGARCH Computer Architecture News, Vol. 13, Issue 3, p.76. (Actually, I suspect that these two are the same thing, i.e. the proceedings of the conference were published as an issue of the journal.) Unfortunately, the paper is only accessible through ACM digital library. http://portal.acm.org/citation.cfm?id=327133&jmp=citings&dl=GUIDE&dl=ACM
Some very random bits of Symbolics docs .
"Symbolics, Inc. is a now defunct computer manufacturer headquartered in Cambridge, Massachusetts, with manufacturing facilities in Chatsworth, California (a suburb of Los Angeles). Symbolics designed and manufactured a line of Lisp machines, single-user computers optimized to run the programming language Lisp. Symbolics also made significant advances in software technology, and offered the premier software development environment of the 1980s and early 1990s. The Lisp Machine was the first commercially available 'workstation' (although that word had not yet been coined)."
Symbolics Lisp Machines, oder: früher war alles
besser [in German]
MP3 and MPEG-4 recordings of a lecture delivered at an event of the infamous Chaos Computer Club e.V.
"Are you tired of buffer overflows? Do you wonder why, when there's a bug, you can't just click on the stack trace, and an editor window with the faulty source pops up - for all parts of the system? Are you sick of losing your uptime because you can't upgrade the kernel without booting?
"Meet the Lisp Machines, a familily [sic] of computers that were running an OS written entirely in Lisp, right down to the hardware. Not only were buffer overflows a problem completely unknown, also the availability of the complete source code at the click of the mouse made for a system of great hack value unmatched by any of the systems available. Remind you, that was at a time before Apple released the Mac.
"This workshop gives an overview of the history of Lisp machines, their implementation principle, tries to look into reasons why such a superior system failed in the market, and gives a demonstration of a running Symbolics MacIvory (if we can make it work again in time)."
European Patent 0113460: Symbolic language data processing
US Patent 4887235. Symbolic language data processing
This patent is a little more kitchen-sink than the other, including the entire microcode. While this reflects the difference in patent legislation between EU and US, it also gives us the opportunity to read the microcode.
3600 Techical Summary.
This "technical" summary is mostly a marketing document, but nonetheless contains some interesting information.
Basic tools for the Lisp programmer.
This document was made available by Rainer Joswig, along with the following comment: "You may wonder whether it still makes sense to get a Symbolics Lisp Machine (for example the Open Genera emulation running on Alpha machines). Genera is the operating system of the Symbolics Lisp Machine and it still has the largest amount of tools for Lisp programming available anywhere. Some of the more basic tools of the Genera operating system for Lisp programmers are explained in this paper. It explains basic editing, inspecting, compiling and debugging Lisp code. Code running on the Lisp Machine is always debuggable—that's a big difference to most of the other systems available."
"This conceptual introduction explains what your Symbolics computer is all about. We urge you to at least skim it before you start daily work with the system. We try here to summarize some of the 'big picture' concepts in your new software environment; knowing about this framework will help you in learning its details more efficiently."
These are quicktime movies demo-ing Genera, made by Rainer Joswig. Unfortunately, I can't find a better link at the moment.
I snarfed the following documents from the net at some point in time, but I couldn't find them there while writing this page, so I provide them here locally. The Genera documentation seems to have disappeared from the place it used to be at. If you think you are the copyright holder to any of these documents and want me to remove them, please drop me an email.
"Symbolics' Genera system represents the accumulation of nearly three and a half decades of software engineering tools, written in several dialects of the Lisp language and several object-oriented extensions to Lisp. Traditionally, Genera has run on a 'Lisp Machine', a line of custom hardware workstations designed specifically to execute the Lisp language. Because of the limited market, this custom hardware has never been able to take advantage of cutting-edge technology such as is available in commodity, RISC-based workstations. At the same time, non-standard hardwares such as the Lisp Machine have fallen into economic disfavor. Nonetheless, Lisp's (and the Lisp Machine's) capability in prototyping, evolutionary problem solving, and super-complex problems has retained a small but dedicated market."
"In response to market pressure to provide Genera's capabilities on commodity hardware, Symbolics chose to implement an 'emulator' that would execute the Lisp Machine instruction set on a standard platform, rather than to port the approximately 1.5 million lines of Genera source code to a single Lisp dialect, as would be necessary to take advantage of a native Lisp compiler on the same platform. It was felt that this approach would result in a shorter time-to-market while still preserving the robustness and flexibility of Genera. At the same time, the emulator approach has meant that the most important features of the Lisp machine in support of Lisp have also been preserved. In particular, the tagged memory architecture supporting dynamic typing, the fast exception handling for complete instruction semantics, and read and write barriers in support of automatic storage management have been preserved."
Xerox to Symbolics conversion
This page, not particularly interesting unless you want to connect a Xerox mouse to a Symbolics or vice versa, seems to be unavoidable when surfing around for LispM information.
Eugene C. Ciccarelli IV: Presentation Based User
Interfaces, AI Technical Report 794.
This is a thesis about Presentation Based User Interfaces; basically, it's some of the early work behind such wonderfulness as the dynamic listener and the display debugger. The text is not Symbolics specific, it's just interesting background reading for the Symbolics enthusiast.
The Texas Instruments (TI) Explorer and MicroExplorer are derived from the LMI machines.
The E3 project
An effort to make an emulator for the TI Explorer (affectionately called Exploder), which was built by Texas Instruments using technology licensed from TI. Contains many references to information about these machines.
TI Explorer Documentation
TI MicroExplorer Documentation
These mostly came out of Xerox, although I think they have also been sold by Siemens. I know relatively little about them; the following links are courtesy of Brian Mastenbrook, who wrote: "I saw your excellent Lisp Machines resource page off of lemonodor, and I don't see anything about the Xerox InterLisp machines at all. These machines have been kind of forgotten over the past decade, due to the (nil) market favor of InterLisp; however, they spawned many important things, not the least being CLOS and PCL. InterLisp-D ran as one system on the Dorado series of Xerox workstations, which were sold re-labled as the 1108. These machines also hosted the famous Xerox Office Environment and the Smalltalk environment. The following papers and links should be a good addition to your collection..."
Some images related to Interlisp
These are from Rainer Joswig, who has done some excellent work making information about various Lisp systems available; by all means, browse around his site.
A Primer to Interlisp-D
More random Interlisp documentation
A zipped quicktime movie showing Brian interacting with
the Xerox environment
The Venue Medley emulator, free for non-commercial
In this section I can't claim too much expertise; there have been a large number of Lisp machine designs not derived from the Knight architecture. Undoubtedly, some of them were brilliant, and as time goes on I would love to extend this section with more information. If you know of any information on the web about these machines, please feel free to contribute. (This holds for documents on other types of Lisp machines too, by the way.)
Sansonnet: M3L -- A Lisp Machine
"The M3L project has been developped [sic] in the LSI laboratory (now called IRIT) at Toulouse Paul Sabatier University in the research team of Professor René Beaufils. Between 1977 and 1982. Jean-Paul Sansonnet and his PhD students: Michel Castan, Christian Percebois, Daniel Botella and Julien Perez worked on a fully integrated Lisp-machine called M3L. This machine had the unique feature of embedding a fully microcoded Lisp interpreter which made it the fastest Lisp machine of its (short) time."
Guy Lewis Steele, Jr. and Gerald Jay Sussman. "Design of LISP-based Processors, or SCHEME: A Dielectric LISP, or Finite Memories Considered Harmful, or LAMBDA: The Ultimate Opcode". MIT AI Lab. AI Lab Memo AIM-514. March 1979. Available online: ps pdf .
An experimental architecture that is based on Lisp as a machine language.
IPSJ JOURNAL Abstract Vol.24 No.05 - 016
Hirotoshi, Hiroshi, Takeshi, Toshio, and Toshifumi. Fast
LISP Machine and List-Evaluation Processor EVAL II :
Processor Architecture and Hardware Configuration.
IPSJ JOURNAL Abstract Vol.24 No.05 - 017
Toshifumi, Toshio, Takeshi, Hirotoshi, Hiroshi: Fast LISP
Machine and List-EvaIuation Processor EVALII : Machine
Evaluation on Interpreter
These two are references to articles about a Japanese Lisp Machine design.
A lot has been written about the rise and fall of Lisp, the Lisp Machine and Symbolics. These are some of the pages you may want to refer to. Some of them are at best tangential, but may be useful in trying to understand the social and business aspects of the Lisp Machine era.
The UNIX-HATERS handbook. A most hilarious book, compiled
from the archives of the UNIX-HATERS mailing list.
Richard P. Gabriel. Lisp: Good News; Bad News; How to Win
Lisp has done quite well over the last ten years: becoming nearly standardized, forming the basis of a commercial sector, achieving excellent performance, having good environments, able to deliver applications. Yet the Lisp community has failed to do as well as it could have. In this paper I look at the successes, the failures, and what to do next.
Eve M. Philips. If it works, it's not AI: a commercial
look at Artificial Intelligence startups. BS/MEng Thesis.
The goal of this thesis is to learn from the successes and failure of a select group of artificial intelligence (AI) firms in bringing their products to market and creating lasting businesses. I have chosen to focus on AI firms from the 1980s in both the hardware and software industries because the flurry of activity during this time makes it particularly interesting.
Graylin, Kjolaas, Loflin and Walker. Symbolics, Inc.: a
failure of heterogeneous engineering.
Another analysis from a business perspective, this one exclusively about Symbolics. ( Comment by Dan Weinreb )
The Lisp Machine: Noble Experiment Or Fabulous Failure?
Draft 11 July 1991 P. T. Withington Symbolics, Inc.
Richard Gabriel. Worse is Better.
This famous paper clearly points out the UNIX/Lisp, New Jersey/MIT, worse-is-better/the-right-thing distinction.
Alan Bawden: PCLSRing: Keeping Process State Modular.
This is the PCLSRing that Gabriel is referring to.