Packing for the flight towards #ICFP23, I wanted to warm up my nix store for a few select projects, when I get “no more disk space“ despite the partition being only 70% full – the dreaded no-more-inodes problem. So now running nix store --gc, and hope I remember to re-enter all relevant nix shells once before the flight. I think #lean ’s (experimental) one-dervation-per-file nix support and me recompiling #mathlib a few times recently is partly to blame…
What to change? - What to change to? - How to change?
Have you ever heard of these questions? Not? Then it is time for another book on your shelf. If you know „The Phoenix Project“, you might like the format of „The goal“ from Eliyahu M. Goldratt.
It’s a super interesting novel about a plant that is in trouble and how it avoids to get closed.
@Phipe La « simplification », au sens du #lean , ça va mal finir… surtout quand on confie la dématérialisation (autre nom de « réduction de postes » ou de « non remplacement des partants ») à des pieds nickelés, alors qu’il y a des gens qui savent faire !
Someone once said that laziness is what kept #Haskell pure, and that's the actually relevant feature. This makes we wonder: will theorem proving languages like #lean, where logical consistency is what keeps them pure, deliver the same elegant experience, while avoiding some downsides of laziness (complex runtime, complicated performance characteristics)?
Getting interested in #Lean upon reading that the community isn't very concerned about whether it meets constructivists' criteria for theoretical soundness