I’m making my own Lisp implementation despite having a tenuous grasp of how Lisp works, because I read 40% of a math paper about Lisp. I’m not even referencing the math paper as I write it I am entirely just making shit up

https://gitlab.com/__kasdeya__/kaslisp/

this is a bad decision and no one can stop me from making it

it genuinely took me more than an entire day to realize that the result of (car (foo bar baz)) should not be foo

also I tried to represent lists like (foo bar baz) as (cons foo (cons bar (cons baz [Python None]))) and it took me a long time to realize why Lisp doesn’t do that (it’s because NIL being an empty list means that it’s still of type list, so you unambiguously know what type it is when you reach it with cdr)

@kasdeya
I hated Lisp and Prolog. Good luck.
Also I realised they wouldn't implement AI.
Industry 1985: We just need a more powerful CPU to do AI.
Me: You can extend memory massively on a Z80 and even have an HDD. If you knew how to do AI, then you could have had slow AI even before 1981 IBM PC.
Never mind 1970s supercomputer mainframes.
Touring Test not an validation of AI, but a parlour game.