View psilo-typeclass.sl
;; Hask-- I mean, psilo with no real typeclasses and one virtual typeclass.
; Inspired by [1], this is an exploration of representing all typeclasses
; through one distinguished class.
;
; This version of psilo does not have "real" typeclasses. However faking them
; with explicit dictionary passing does seem to correctly infer and check the
; constraints.
;
; The goal then is to figure out how best to represent "real" typeclasses as a
; distinguished feature in the language implementation given that we are
View io.hs
{-# LANGUAGE RankNTypes #-}
import Prelude hiding (IO, getLine)
import qualified Prelude as P
import System.IO.Unsafe
-- * The Foreign Function Interface
-- | FFI values permit interfacing with foreign functions, such as low-level IO
-- operations, memory management operations, or bindings to other user-level
View install-arch-linux-rpi-zero-w.sh
#!/bin/sh -exu
dev=$1
cd $(mktemp -d)
function umountboot {
umount boot || true
umount root || true
}
# RPi1/Zero (armv6h):
View free-compiler.hs
{-|
- Example of using free constructions to build a flexible little compiler.
-
- The goal here is not necessarily efficiency but readability and flexibility.
-
- The language grammar is represented by an ADT; however, instead of
- recursively referring to itself it instead references a type variable.
-
- We derive instances of 'Functor' and 'Traversable' for this type.
-
View index.html
<!DOCTYPE html>
<html>
<head>
<title>Large Type</title>
<link rel="stylesheet/less" type="text/css" href="style.less">
<script type="text/javascript" src="lib/jquery-2.1.4.min.js"></script>
<script type="text/javascript" src="lib/less.min.js"></script>
</head>
<body>
<div class="out"><span contenteditable autofocus>*hello*</span></div>
View yet-another-conway.hs
{-# LANGUAGE DeriveFunctor #-}
import Control.Applicative
import Control.Comonad
import Control.Monad (forM_)
-- These two typeclasses probably make some people groan.
class LeftRight t where
left :: t a -> t a
View markov.py
###
# Markov models!
#
# A Markov model of some process is a model of different states you can be in,
# where each state links to other states. These links are weighted by some
# probability score. Markov models are designed to model systems where future
# states depend only on the present state.
#
# In the case of words you may define "present state" to be "the current word,"
# or perhaps "the last 3 words."
View audio
#!/bin/bash
###
# Script to launch jackd and reroute PulseAudio through it. And optionally
# start fluidsynth.
#
# READ THE NOTES PLEASE
###
# Usage (assuming this script is named `audio`, is in your PATH, and is
View output.txt
Code: (+ 5 5)
Free (AApp (Free (ASymbol "+")) [Free (ANumber 5),Free (ANumber 5)])
---
Code: (lambda (x) (* x x))
Free (ALambda [Free (ASymbol "x")] (Free (AApp (Free (ASymbol "*")) [Free (ASymbol "x"),Free (ASymbol "x")])))
---
Code: ((\ (x) (* x x)) 5 (+ 10 2))
Free (AApp (Free (ALambda [Free (ASymbol "x")] (Free (AApp (Free (ASymbol "*")) [Free (ASymbol "x"),Free (ASymbol "x")])))) [Free (ANumber 5),Free (AApp (Free (ASymbol "+")) [Free (ANumber 10),Free (ANumber 2)])])
---
Code: '(1 2 3)
View pid.lhs
PID controller in Haskell
===
A major project I want to embark on at some point in the future is making a
quadrotor. I've made one before but I was at the mercy of a lot of off-the-shelf
software that I'm not altogether sure was entirely correct, either.
So I want to eventually program a quadrotor (or similarly awesome robot thing)
and I would really enjoy doing it in Haskell. It has a low footprint and is
already used in other real time settings.