Lisp Machine Question (and some geekery). 
Author Message
 Lisp Machine Question (and some geekery).


>    Section 2-20 of the Lisp FAQ says the following ...
>It has left
>me with the desire to use MACH to create a Virtual Lisp Machine, sitting
>on top of MACH, with a bcode interpreter which more or less mimics the
>3600 (or something).  Has anyone ever tried this?  None of the manuals
>I could find had a complete 3600 Instruction set.  Does anyone know if such
>a thing existed (external to Symbolics, at any rate)?

Yes. Before the original MIT CADR (lispmachine) hardware was built
there was an implementation of the 16-bit instruction set, written
in PDP-10 assembler language. This was used to debug the compiler
and runtime system and to bootstrap the actual machine.

The CADR instruction set was actually for a Scheme Machine, with elegant
support for tail recursion. Symbolics threw out this support when they
made various "speed" improvements to the procedure call sequence.

I would recommend you attempt to obtain the CADR document from somebody at
the MIT AI laboratory. Then talk to John Preston at the MIT patent office
about the possibility of obtaining the full sources to System 99, which was
circa 1985, the last MIT release of the full CADR operating system, including
compiler, network software, windowing software etc.

There was also sources left over from the final demise of Gigamos. The LMI-LAMBDA
instruction set was the same as the CADR (and the TI explorer for that matter),
with some slight enhancements. The LMI sources included a working realtime
garbage collector and TCP-IP implementation and improved filesystem.
But you couldn't use these until you had the license from MIT.

MIT was at one time restricted by an agreement with Symbolics to charge a huge
sum of money for the MIT sources, even as the sources became less and less
valuable commercially. But with the demise of the specialized workstation
market and it being 1996 and all, perhaps things have changed.

The 16-bit CADR instruction set has a lot of advantages over 8-bit bytecoding.

On the other hand, you might instead consider the techniques described
in Gordon Bell's book on computer engineering "A DIGITAL PERSPECTIVE"

Sun, 19 Jul 1998 03:00:00 GMT  
 [ 1 post ] 

 Relevant Pages 

1. Lisp Machine Question (and some geekery).

2. Lisp Based Machine Code (Simulators) [was Re: Lisp Machines]

3. Lisp Based Machine Code (Simulators) [was Re: Lisp Machines]

4. Questions about Symbolics lisp machines

5. obscure old Lisp machine question...

6. Lisp Machines and Lisp on a Chip

7. Lisp Machine architecture and other hardware support for Lisp

8. Ideas for new LISP Machines

9. Interpreted code penalties (was: Re: Ideas for new LISP Machines)

10. Free Symbolics Lisp Machines

11. A cool device that could make a lisp machine on modern hardware

12. Parallel Lisps for Parallel Machines


Powered by phpBB® Forum Software