miniblog.

Dynamically types languages have the advantage of concrete values, so static languages *must* have helpful errors:

ripgrep is even faster than ag! https://blog.burntsushi.net/ripgrep/ Exciting! Rust (plus the talent of those involved) is preventing Wirth's law.

It is terrifying to learn that all Intel CPUs contain a separate system with full access and a network connection:

The MIT Lisp Machine manual argues in favour of allowing users to patch internal details of classes:

Interestingly, the FSF's biggest objection to GitHub is not the proprietary backend! https://www.gnu.org/software/repo-criteria-evaluation.en.html (it's the JS and censorship)

Emacs command of the day: ivy-read-action, bound to C-M-a when ivy is active. Change the action on RET before you choose a candidate!

I'm familiar with some pretty arcane Emacs keybindings, but I'm really impressed with C-φ in lispy!

Every time that I play with Pharo (Smalltalk), I miss something when I go back to my usual language tooling.

Viewing a codebase as an immutable, content addressable tree: https://unisonweb.org/2016-09-17/editing.html (enables some really interesting features!)

Sorted copyright assignment for @guilelang! Hurrah!

TIL there's a service on freenode that conveniently lists channels:

Emacs command of the day: crux-sudo-edit from https://github.com/bbatsov/crux . Great for editing the current file, if it's owned by root.

I've just discovered I can double performance of my elisp project by let-binding gc-cons-percentage to 0.8! Woohoo!

Apparently function calls are cheap in elisp. I rewrote a recursive function as iterative and was pleasantly surprised that perf matched.

Although 'time travelling debuggers' sound cool, I would find them easier to explain if we simply called them 'debuggers with undo'.