Home
Login Register Member List
PDP Planet Forums   » PDP Planet Forums   » Software   » LISPs that run on Tops-10, Top...  

Welcome Guest ( Login or Register )

LISPs that run on Tops-10, Tops-20
Author
Message
Posted 6/3/2005 2:49 PM
Forum Newbie
Forum Newbie

Group: Forum Members
Active: 8/3/2005
Posts: 2
(I started this conversation by email; Rich Alderson suggested moving it to this forum.)

I've been doing some work to collect source code, documents, etc. for various LISP systems (see http://community.computerhistory.org/scc/projects/LISP/), and it would be great if people could try out some of these. So I'm curious which LISPs could be made to run on one of the restored machines. Rich notes "We have MACLISP on the Toad-1."

Here are the possibities I've run across:

MacLisp
Phil Budne. Source files and text documents.
ftp://ftp.ultimate.com/pdp10/maclisp/
* These items were obtained from a TOPS-20 system at XKL.com:
o MACLISP LISP sources including COMPLR.LSP lisp.tar.gz
o MACLISP .DOC and .MID files including LISP-NEWS.DOC from April 28, 1981. lisp2.tar.gz
* These items were obtained from tape images of Chemistry Department disk packs from the
Stevens Tech TOPS-10 timesharing system:
o MIDAS sources for building MACLISP version 804 and MIDAS. Circa June 1979. maclsp804.tar.gz
o MACLISP version 861 (no additional MIDAS files) and Maclisp Reference Manual, March 3, 1979.
maclsp861.tar.gz

Stanford Lisp 1.6 for TOPS-10:
Lynn H. Quan and Whitfield Diffie et al. LISP 1.6. DECUS Library Program 10-59, May 1, 1973.
http://pdp-10.trailing-edge.com/decuslib10-01/index.html

also:

University of California, Irvine. UCI-LISP. DECUS Library Program 10-210, March 27, 1975. Online at pdp-10.trailing-edge.com
http://pdp-10.trailing-edge.com/decuslib10-04/index.html


Interlisp on TOPS-20:
http://pdp-10.trailing-edge.com/decuslib20-01/index.html


Standard Lisp
I.B. Frick. Manual for Standard LISP on the DECSYSTEM 10 and 20. Technical Report TR-2, Utah Symbolic Computation Group, University of Utah, July 1978.


Others?
IP Logged
This member is offline.
Posted 6/6/2005 5:50 PM
Forum Newbie
Forum Newbie

Group: Forum Members
Active: 7/7/2005
Posts: 2
Chuck Hedrick wrote an implementation of Common Lisp for TOPS-20
in the early 80's, contemporaneously with the Common Lisp design
effort. As far as I know, it was the only implementation of Common
Lisp for the PDP-10 architecture, and required extended addressing.

CMU ran a hacked-up version of MacLisp that might be interesting
to find, as I know that a version of Terry Winograd's SHRDLU
natural language understanding system survives that was ported at
CMU to the local MacLisp flavor.

--Bill
IP Logged
This member is offline.
Posted 6/8/2005 8:44 AM
Forum Newbie
Forum Newbie

Group: Forum Members
Active: 8/3/2005
Posts: 2
Bill,

Thanks very much for mentioning that. I looked up Hedrick via Google, and ran across this fascinating comp.lang.lisp posting from 1994 with email from 1981-1984 by the Common Lisp "inner circle":

http://groups-beta.google.com/group/comp.lang.lisp/msg/0ffe52efb6bd479b

Also, Steele and Gabriel in their 1993 HOPL paper (http://www.dreamsongs.com/NewFiles/HOPL2-Uncut.pdf) say:

"By the middle of the 1970’s it became apparent that the 18-bit address space of the PDP-10 would not provide enough working space for AI programs. The PDP-10 line of computers (KL-10’s and DEC-20’s) was altered to permit an extended addressing scheme, in which multiple 18-bit address spaces could be addressed by indexing relative to 30-bit base registers.

However, this addition was not a smooth expansion to the architecture as far as the Lisp implementor was concerned; the change from two pointers per word to only one pointer per word required a complete redesign of nearly all internal data structures. Only two Lisps were implemented for extended addressing: ELISP by Charles Hedrick at Rutgers [?]and PSL."

Any idea where to track down source and object code for ELISP?

Also, do you have any suggestions for who to contact re CMU's hacked-up MacLisp?

Paul
IP Logged
This member is offline.
Posted 6/13/2005 9:24 AM
Forum Newbie
Forum Newbie

Group: Forum Members
Active: 6/23/2005
Posts: 1
I think that MDL qualifies as LISP. Time must have faded arguments to the contrary.

I don't know what hope there is of finding it, though. The ZORK enthusiasts would have by now if it could be.
IP Logged
This member is offline.
Posted 7/7/2005 8:36 AM
Forum Newbie
Forum Newbie

Group: Forum Members
Active: 7/7/2005
Posts: 2
The version of SHRDLU I was thinking of is the one below:

ftp://ftp.cs.cmu.edu/user/ai/areas/classics/shrdlu/

Reading the notes, it appears that the "hack" was simply
the porting of MAClisp itself from ITS to TOPS-10 at CMU.

It turns out that another, probably more original version
is available from Terry Winograd's site at:

http://hci.stanford.edu/~winograd/shrdlu/

David Touretsky wrote the "pocket guide" to MAClisp that
everybody at CMU used, and I believe he maintained the local codebase,
though I'm not sure about that. He's a professor at Princeton
now but he might have kept an old tape or have some war stories
to tell.

I don't have any further information on ELISP, but Charles
is still at Rutgers and on the web:

http://toolbox.rutgers.edu/~hedrick/

The guy to talk to about anything related to the PDP-10's
at CMU would be Howard Wactlar (wactlar@cmu.edu). He was
the director of research computing in the CS department when
I was at CMU, and I see he's moved up to Vice Provost these
days, but still running research computing. Howard was a
big fan of 10's. We ran a modified version of TOPS-10, and
I believe Howard did much/most of the hacking. Sometime around
1981 or so, he found a used KI somewhere, and had it installed
in the most visible location in the machine room. We were still
running a few KA's at the time, but it still seemed a bit odd to
be installing a machine as old as the KI at a time when we had
a 2060 and were buying VAXen like crazy. Some folks
I knew who were better connected in CS (I was a mere undergraduate
part-time employee) called that machine "Howard's baby". It was
a great-looking machine -- I wonder what ever became of it.

--Bill
IP Logged
This member is offline.
Posted 7/19/2005 12:04 PM
Forum Newbie
Forum Newbie

Group: Forum Members
Active: 7/19/2005
Posts: 1
Quote: Chuck Hedrick wrote an implementation of Common Lisp for TOPS-20in the early 80's, contemporaneously with the Common Lisp designeffort. As far as I know, it was the only implementation of CommonLisp ...

I have a copy of the Common Lisp implementation, which we used for many years in the CS education at Uppsala University. We did a few improvements and bug fixes - I'm not sure I still have the original, but surely our hacked version. I know Systems Concepts also worked on improving it; perhaps also XKL did?

IP Logged
This member is offline.
Posted 7/19/2005 12:07 PM
Forum Member
Forum Member

Group: Forum Members
Active: 8/6/2012
Posts: 10
I can state categorically that XKL did nothing with Common Lisp for Tops-20.
Rich Alderson (RichA@Vulcan.com)
Not speaking as the forum moderator
IP Logged
This member is offline.
Posted 1/11/2006 10:50 PM
Forum Newbie
Forum Newbie

Group: Forum Members
Active: 1/20/2006
Posts: 4
I believe at WPI (Worcester Polytechnic Institute my alma matter) they ran MacLisp - I may still have the manual.

IP Logged
This member is offline.
Showing page 1 of 1 - 8 messages found
Next Newest »

Powered by InstantForum.NET v3.4
Copyright © 2002, 2004, InstantASP Ltd. All rights reserved.