Don't know if my language is considered Lil' enough for this, but it's a pure, lazy functional language based upon Miranda (progenitor language to Haskell) that compiles to x86-64 asm. ~6700 SLOC for the (self-hosted!) compiler, and ~3300 SLOC additional for the extensive library of functional data structures and functions.
I really loved Miranda back when I learned about it. I still have the book. I think it never took off because it was quite expensive for universities to use. Im sure David Turner regrets his price model today. Now he has made Miranda available here https://www.cs.kent.ac.uk/people/staff/dat/miranda/
Yes, the open-source release he did is what introduced me to Miranda. I rewrote a lot of my previous Haskell solutions to Advent of Code puzzles with it, and liked it so much I decided to try to improve on it ;-)
That's what led to Admiran. I originally wrote Admiran in Miranda, then bootstrapped from that to self-hosting when it was stable enough to do so. The original Miranda combinator compiler / interpreter took 20 minutes to compile all of Admiran, while the self-hosted version now takes 20 seconds.
One of the grad students of David Turner has taken up maintenance on the original Miranda source; the repository is now at https://codeberg.org/DATurner/miranda
My little language Newt is 7 kloc. Dunno if it's worth including, it's mostly an exercise to learn how these things work and is not as polished as I'd like.
- Self-hosted
- Compiles to javascript
- Bidirectional typechecking with NbE (based on elaboration zoo)
- Dependent type checking
- type classes
- ADTs with dependent pattern matching
- TCO (trampoline for mutually tail recursive functions)
- Erasure of compile-time only values (0, ω quantities, but not linear)
Either newt was already in the list, or it got added. We talked a bit about using our languages for AoC 2024 -- looks like you've been keeping busy working on it!
Yeah it has been fun. Lots of directions I can take it:
Since I have an LSP, I've got faster turn around and can add editor functionality that requires poking at the compile state. That's my current thread.
I have a C backend on hold, while I think about how I want to represent data without boxing everything and about whether I want to do reference counting or GC. (Reference counting unlocks "counting immutable beans" if I decide to give that a go, but I'd also like to try implementing GC someday.)
I should do some browser interop stuff and write something other than a compiler in my language.
And there are language enhancements: implementing "Do unchained" from Lean, automatic handling of lazy and/or async modalities, deriving implementations of classes, ...
Reminds me of Ghuloum's "An Incremental Approach to Compiler Construction" [1]. The tutorial [2], takes you through making a scheme compiler. It starts by compiling an integer to x86 assembly, using the output of gcc's assembly as reference. Then it goes through to tail calls and heap allocation, with a working compiler at every step. Tests at [3]. Mentioned in [4,5]. Gholoum went on to write Ikarus scheme which continued that work [6], but it never moved to 64 bit and the last release was in 2008.
I've been playing with my own Smalltalk implementation. I guess it qualifies as a tiny _language_, but the class library is huge!
Yes, I know that all who do not understand Smalltalk are doomed to re-implement it, poorly. I'm just cutting out the middleman - and this allows it to do things Smalltalk normally doesn't. It allows me to think so far outside the box it's not even visible anymore.
I've always found OCaml's (* ... *) comments annoying, because it needs shift for both characters. But I suspect it's easier to type on a french keyboard.
Keyboards of early MIT systems[1] and Lisp Machines[2] had brackets (parentheses) and square brackets on the same key where square brackets and curly brackets are on modern keyboards.
Wow, this is so cool! Very overwhelming with so many interesting things at once. I wish I had a few months with nothing else so I could surround myself with only this. :)
The hardest part with small languages isn't the parser, it's the standard library and error messages. Getting a helpful IDE experience in that footprint is a significant engineering challenge.
Don't know if my language is considered Lil' enough for this, but it's a pure, lazy functional language based upon Miranda (progenitor language to Haskell) that compiles to x86-64 asm. ~6700 SLOC for the (self-hosted!) compiler, and ~3300 SLOC additional for the extensive library of functional data structures and functions.
https://github.com/taolson/Admiran
I really loved Miranda back when I learned about it. I still have the book. I think it never took off because it was quite expensive for universities to use. Im sure David Turner regrets his price model today. Now he has made Miranda available here https://www.cs.kent.ac.uk/people/staff/dat/miranda/
Yes, the open-source release he did is what introduced me to Miranda. I rewrote a lot of my previous Haskell solutions to Advent of Code puzzles with it, and liked it so much I decided to try to improve on it ;-)
That's what led to Admiran. I originally wrote Admiran in Miranda, then bootstrapped from that to self-hosting when it was stable enough to do so. The original Miranda combinator compiler / interpreter took 20 minutes to compile all of Admiran, while the self-hosted version now takes 20 seconds.
One of the grad students of David Turner has taken up maintenance on the original Miranda source; the repository is now at https://codeberg.org/DATurner/miranda
My little language Newt is 7 kloc. Dunno if it's worth including, it's mostly an exercise to learn how these things work and is not as polished as I'd like.
- Self-hosted
- Compiles to javascript
- Bidirectional typechecking with NbE (based on elaboration zoo)
- Dependent type checking
- type classes
- ADTs with dependent pattern matching
- TCO (trampoline for mutually tail recursive functions)
- Erasure of compile-time only values (0, ω quantities, but not linear)
- Web playground
- LSP (added this month)
- Syntax is similar to Agda / Idris / Haskell
https://github.com/dunhamsteve/newt
Either newt was already in the list, or it got added. We talked a bit about using our languages for AoC 2024 -- looks like you've been keeping busy working on it!
Yeah it has been fun. Lots of directions I can take it:
Since I have an LSP, I've got faster turn around and can add editor functionality that requires poking at the compile state. That's my current thread.
I have a C backend on hold, while I think about how I want to represent data without boxing everything and about whether I want to do reference counting or GC. (Reference counting unlocks "counting immutable beans" if I decide to give that a go, but I'd also like to try implementing GC someday.)
I should do some browser interop stuff and write something other than a compiler in my language.
And there are language enhancements: implementing "Do unchained" from Lean, automatic handling of lazy and/or async modalities, deriving implementations of classes, ...
Reminds me of Ghuloum's "An Incremental Approach to Compiler Construction" [1]. The tutorial [2], takes you through making a scheme compiler. It starts by compiling an integer to x86 assembly, using the output of gcc's assembly as reference. Then it goes through to tail calls and heap allocation, with a working compiler at every step. Tests at [3]. Mentioned in [4,5]. Gholoum went on to write Ikarus scheme which continued that work [6], but it never moved to 64 bit and the last release was in 2008.
[1] http://scheme2006.cs.uchicago.edu/11-ghuloum.pdf
[2] https://web.archive.org/web/20061217101846/http://www.cs.ind...
[3] https://web.archive.org/web/20090618030856/http://www.cs.ind...
[4] http://lambda-the-ultimate.org/node/1752
[5] https://news.ycombinator.com/item?id=21627615
[6] https://web.archive.org/web/20090928074841/http://ikarus-sch...
I've been playing with my own Smalltalk implementation. I guess it qualifies as a tiny _language_, but the class library is huge!
Yes, I know that all who do not understand Smalltalk are doomed to re-implement it, poorly. I'm just cutting out the middleman - and this allows it to do things Smalltalk normally doesn't. It allows me to think so far outside the box it's not even visible anymore.
loon is a lisp! https://github.com/ecto/loon
idk how I haven't crossed a lisp with square brackets but dang I am sorta stunned at how I've never even envisioned it? thanks
No shift key needed for square brackets!
Curl was a proprietary Lisp that {curly brackets} and was designed in the 1990s to build web applications.
https://en.wikipedia.org/wiki/Curl_(programming_language)
I've always found OCaml's (* ... *) comments annoying, because it needs shift for both characters. But I suspect it's easier to type on a french keyboard.
Knuth solves the bracket issue by redefining his keymap to swap () with [] and + with = (macos keymap files found at the bottom of this page: https://www-cs-faculty.stanford.edu/~knuth/programs.html )
In high school I tried spec'ing a lang that excluded shift key usage... fun times.
Keyboards of early MIT systems[1] and Lisp Machines[2] had brackets (parentheses) and square brackets on the same key where square brackets and curly brackets are on modern keyboards.
1: http://xahlee.info/kbd/sail_keyboard.html
2: http://xahlee.info/kbd/space-cadet_keyboard.html
There's a few lispy languages that support using square brackets as an alternative to parens. Racket even has curly braces too.
The above really reminds me of tcl, though, which uses square brackets to force evaluation of the command they wrap.
Wow, this is so cool! Very overwhelming with so many interesting things at once. I wish I had a few months with nothing else so I could surround myself with only this. :)
Fluent – 4K lines – including parser, interpreter, standard library, IDE, UI, docs, examples. Will grow though.
https://github.com/mlajtos/fluent/blob/main/client.tsx
I'll add it! Thanks.
EDIT: Actually, it's not quite "ML-family" enough for this post. But it is a remarkably cool project! :)
If you're accepting additions here is a fun one some friends and I did as experiment at the Topos Institute: https://github.com/ToposInstitute/polytt
And here is a set of single file lambda calculus implementations with a variety of extensions: https://github.com/solomon-b/lambda-calculus-hs
`polytt` is kind of an ended experiment but that lambda calculus repo i plan to extend in the near future.
> Actually, it's not quite "ML-family" enough for this post. But it is a remarkably cool project! :)
Fair. Thank you :)
fun collection. would love your comment on roxlang(https://roxlang.com/)
disclaimer: I'm the author
Another crazy one is SectorLISP, 223 lines of asm
https://justine.lol/sectorlisp2/
The hardest part with small languages isn't the parser, it's the standard library and error messages. Getting a helpful IDE experience in that footprint is a significant engineering challenge.
[dead]