Tootfinder

I have written another essay for my blog. This one is about where I'm up to with my mad #Lisp system, and is *really* obscure!
https://www.journeyman.cc/blog/posts-outpu

Vector space, Pages, Mark-but-don't-sweep, and the world's slowest ever rapid prototype
...it was intended, largely, to be a giant thought experiment. But now enough of it *does* work that I can see fundamental design mistakes, and I'm thinking about whether it's time to treat it as a rapid prototype...

In #Clojure, if you query a set for a member, and that member is present, that member is returned:
user=> (#{:a :b :c} :a)
:a
The traditional #Lisp function ASSOC has the signature
(ASSOC store key) => value
where store is assumed to be a list of (key . value) dotted pai…

Two days ago, I had eight (out of one hundred and forty) unit tests failing, and I wasn't happy with my code. Last night, I had forty failing, and I thought things were improving. This morning I have one hundred and twenty four failing, so I *know* I'm on the right path!
#Software
#Lisp

Right, problems for today. First. #Lisp problems. I think the bug in `cond` fires when a clause succeeds but returns `nil`.
First, write a unit test which checks for that, but run that test on my laptop where it physically cannot generate millions of stack frames.
Second, rewrite `cond` to call a separate helper function, `cond_clause`, which takes one arg and returns `nil` on failure, `…

Woohoo! I just typed
`(mapcar (lambda (n) (:documentation (meta n))) (keys (oblist)))`
into the REPL of my frankenlisp, and printed out the documentation of every function it knows.
I am getting close to the point that other people can play with this...
#Lisp

Last night I discovered a huge 'new' bug in the integer arithmetic of my #Lisp system which I'd been unaware of because I don't have a unit test for it and it isn't something I'd been worried about and don't normally do.
Weirdly, it makes me feel hopeful. It is this:
:: (- 5 4)
1
:: (- 4 5)
4,294,967,295
What's happening here? Integer…