miniblog.

Read-Eval-Print-λove v003 - Forth is live https://us7.campaign-archive2.com/?u=9e5111743fcbf47abc854fcd5&id=9a549ee516&e=8674ae8035 (interactive programming, little languages, highly recommended!

TIL there's a proven-correct compiler for standard ML! https://cakeml.org/ This is an incredible feat, and uses validated ISA models.

Accessing individual bits in floats using constexpr https://brnz.org/hbr/?p=1518 (rather tricky, but a great exploration of float representation)

Playing flappy bird in Mario using bugs for arbitrary code execution https://www.youtube.com/watch?v=hB6eY73sLV0 (incredible, and done by hand!)

I've discovered https://github.com/gdbinit/Gdbinit as a brilliant default gdb config. Tons of useful information, colour coded!

The smallest piece of great Pharo code https://medium.com/concerning-pharo/tiny-yet-so-beautiful-1ef5149c910e (explores adding methods to built-in classes and idiomatic Smalltalk design)

A rather elegant DSL for assertions in Smalltalk: https://dionisiydk.blogspot.co.uk/2016/03/new-version-of-statespecs-20.html (via @chaetal)

€5 lisp machine with real time OS!

I miss Hoogle when I work with other strongly typed languages. So many of my questions are "I'm looking for a function with this type".

Chicago police has a Minority Report style system that tries to predict who will commit crimes: https://www.theverge.com/2014/2/19/5419854/the-minority-report-this-computer-predicts-crime-but-is-it-racist (false positives?)

Why coroutines won’t work on the web https://calculist.org/blog/2011/12/14/why-coroutines-wont-work-on-the-web/ (you need to reason about preemption!)

Awesome Emacs package of the day: firestarter https://github.com/wasamasa/firestarter — great for experiments e.g. -*- firestarter: (compile "clang %f") -*-

ASM programming is hard work. It's fun and super educational, but it's slow going (partly complexity and partly familiarity I suspect).

Another reason to move to Python 3: a useful repr() for regexp match objects! Example: <_sre.SRE_Match object; span=(0, 3), match='foo'>

I've stopped considering my email address as private data. Between GitHub and mailing lists it's too easy to obtain.