Hacker News new | past | comments | ask | show | jobs | submit login
Let’s LISP like it’s 1959 [video] (youtube.com)
183 points by DyslexicAtheist on Feb 28, 2019 | hide | past | favorite | 22 comments



I'm currently working through the implementation of my own personal Lisp, following the awesome MAL process [0]. It's instructive to compare the result with the language as defined in the early 60s and available to programmers then (e.g. as defined by the Lisp manual referenced in the video). One thing that immediately stood out was the use of GOTO ('GO' actually) in the early Lisp rather than some sort of loop or tail recursion.

Incidentally, I've been wondering what app I should write to test my Lisp. The obvious end-state mentioned in the MAL guide is to self-host a Lisp, but I'm thinking it would be more appropriate (and fun) to re-implement one of the early chatbots like ELIZA [1].

[0] https://github.com/kanaka/mal/blob/master/process/guide.md

[1] https://en.wikipedia.org/wiki/ELIZA


The book Paradigms of AI Programming is available for free on github, and implements something like ELIZA in Common Lisp.

I really recommend it.

https://github.com/norvig/paip-lisp/blob/master/README.md


Great recommendation, thanks


Waiting for it and can it port to multiple platform like pi, iOS, Andorid, browser ... etc. Obviously can be a joint effort but once the basic is done, we can have a new and refreshing effort.

I always think that why PAIP is great you always have this feeling like Star Trek NG. Suddenly you have a function that works. BTW, I think the other web sites http://www.gigamonkeys.com/book/ would be good.

Not one person but if there is a basis we can grow the community, especially if one can port to Pi and iOS.


I had fun doing last year's https://adventofcode.com in my own Lisp dialect. It's a pretty big stress for a personal toy language, but not too out there.

(Added: the code: https://github.com/darius/squeam/blob/master/eg/advent-of-co...)


I would consider writing a small standardlib for your lisp. While writing it you will encounter things you can optimize perhaps.


Cool! That is a great offer. If you watch my progress SITREPs at [0] you'll get a feel for when it might ready. I'm behind on the SITREP updates but I'm in fact just debugging macro expansion (step 8) at the moment.

[0] https://www.non-kinetic-effects.co.uk/blog/2019/01/21/MAL-2


i read his comment as “[if i were you] i would consider...”.


Yes, I'm sorry. It was meant as an answer to 'Incidentally, I've been wondering what app I should write to test my Lisp.'


My bad! And I can't even claim not to be a native-English speaker.

It is a good idea and in fact I'm already sort of going down that route. I'm finding that using macros to define new functionality (rather than extend the core set of built-in functions) is a good way to tease out implementation issues as they somewhat stress the underlying mechanisms.

I think my first lib functions will be IO-related.


I too might have a couple things in mind. How do I get in touch with you (gp, KineticLensman although happy to hear from immediate parent too)?


If you're interested, it's also possible to run the actual original Lisp 1.5 on emulated hardware via SIMH.

See this page: http://web.sonoma.edu/users/l/luvisi/

I got it running on my Mac a few months ago. SIMH is on Homebrew.



Let's LISP again, like we did last summer!


The things we Lisped last summer, we'll remember AI winter long?


he is making a Lisp, he is checking it twice ...


this was an excellent history lesson on multiple levels. The final personal remarks about how the speaker connects the ethos of open source to his family history was an unexpected turn.


Is Steve Russell still on hand to demonstrate the PDP-1 at the Computer History Museum in Mountain View from time to time?


He and Peter Samson gave the Spacewar demo last August 4th, on the Vintage Computer Festival day.


If you’re interested in hacking on some modern Common Lisp code, Rigetti Computing (which sells access to quantum computers) open-sourced their SDK, which includes a high-performance quantum simulator called the QVM [1] and an optimizing compiler quilc [2] for their quantum programming language called Quil. All of it is written in CL.

[1] https://github.com/rigetti/qvm

[2] https://github.com/rigetti/quilc


There seems to be some attraction to CL for QC researchers. I also used CL for my PhD's QC compiler and VM. Although my low-lv QVM was written in C, it takes s-expr input. I'm a bit sad it wasn't cited by the Quil guys :p

https://github.com/yvdriess/mcc

https://github.com/yvdriess/qvm


Anyone that likes Lisp should take a look at Factor. Even simpler syntax (RPN instead of prefix, less parenthesis), Lispy generics, functional, procedural, scripting, arbitrary data literals, full of meta goodness, clean good small understandable library, batteries included, sub-languages, REPL + image + source files, low and high level. Inspired in Lisp, Joy, Smalltalk and Forth (and with some things in common with Rebol).




Consider applying for YC's Summer 2025 batch! Applications are open till May 13

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: