Pascal Bourguignon recreates machine-readable source for LISP 1.5

Pascal Bourguignon encountered this item on my History of LISP web site:

  • LISP system assembly listing. “FIELD TEST ASSEMBLY OF LISP 1.5 SEPTEMBER 1961”, labeled “Bonnie’s Birthday Assembly”. M.I.T. Museum, donated by Timothy P. Hart and scanned by Jack Harper. PDF (16MB)

and promptly began reconstructing machine-readable source. This morning he announced on comp.lang.lisp his progress (he’s typed in the source, patched it and Dave Pitts’ assembler to nearly recreate the listing, and is close to running it on the emulator). As he says in a README file of his distribution:

This card deck can be assembled with asm7090-2.1.4 applying the small patch ‘asm7090.patch’ to get a listing as identical as possible. asm7090 prints ‘0’ in the generated words for symbols under different headers, so we cannot make a complete word-for-word comparison of the generated code from the listing, until we modify asm7090 in this respect.

The objective is to recover a perforation for performation image of the Source. The same columns, the same typoes should be reproduced.

If you’d like to help Pascal find the remaining errors, or have the LISP 1.5 compiler sources or LISP 1.5 application sources, you can contact Pascal at the email address in the above-mentioned README file. Please also send me email or post a comment to this entry!

Update 2014/05/10: =>, etc.; 2016/1/2: =>!topic/comp.lang.lisp/Z7HKvfJxhww]

Update 2024/05/08: actually fixed the comp.lang.lisp link, and changed distribution link to

3 thoughts on “Pascal Bourguignon recreates machine-readable source for LISP 1.5”

  1. I have a goal related to this one. I would like to adapt the original code to the modern CPU. My goal is to duplicate the relationship between the Symbolic Expression and the Machine in an effort to understand the original intent as reflected in McCarthy’s Recursive Functions of Symbolic Expressions. In order to do that, I must start here: understanding the original implementation and the machine it ran on.

  2. That is the object file, so in principle you would feed it to a 7090 simulator. I gather from Pascal’s comp.lang.lisp posting that he didn’t try that because he knew he hadn’t eliminated all the typos from his source file, lisp15.src. I haven’t heard any more from Pascal since June; you might contact him directly.

Comments are closed.