Sabitlenmiş Tweet
can't find a topic for your next programming languages paper? worry not! type system idea generator: type.systems
English
joomy
6K posts
joomy
@joomy
researcher @Bloomberg. 📜 somehow a computer doctor. tweets about formal verification, compilers, types, metaprogramming, linguistics. 🇹🇷
New York City Katılım Şubat 2010
966 Takip Edilen3.9K Takipçiler
@abecedarius @rsnous @tobyshooters it was one of my inspirations! (and we cited it in the paper: joomy.korkutblech.com/papers/kip.pdf)
English
@rsnous @joomy @tobyshooters reminds me of metacpan.org/dist/Lingua-Ro… (though I don't actually know any such language, natural or artificial)
English
saw @joomy talk about his Turkish-like morphosyntactic programming language github.com/kip-dili/kip where arguments can go in any order (because inflection, not word order, carries the meaning)
& reminded me of @tobyshooters on geometric code layout (critique of code formatters)
cristóbal@tobyshooters
@pelavarre @rsnous The secret is that this whole endeavor is just me trying to find geometric code layout that dies when you use a formatter. 4.1 was where I started!
English
joomy retweetledi
During today's #LangSec Workshop at #IEEESP2026 (1:15 PM PDT), @Joomy Korkut delivers a keynote outlining the design, implementation challenges, and early lessons learned in integrating verified programs into real-world C++ systems with Crane
bloom.bg/4wCNU3i
#opensource
English
"... In this paper we introduce temporal separation logic."
charlie ☆@asterhound
this idiom confuses me so much wdym i cant have my cake and eat it too? yes i can?? its literally my cake??
English
joomy retweetledi
@kirancodes Lean has GHC or SYB style Generics? my impression was that that kind of generics does not mesh well with dependent types.
English
There's one correct answer and it's Lean
One Happy Fellow@onehappyfellow
which statically typed programming language does generic programming (as in Data.Generic in Haskell) and macros (as in lispy macros or Template Haskell)?
English
my C++ surprise of today: you thought smart pointers would solve your memory management problems? it turns out, if you have a recursive data structure, and a big enough instance of it, the default destructor can cause a stack overflow! (so you need a custom iterative destructor)
English
@joomy @peter0x44 Description of this thing.. contradictory statements
English
@joseph_h_garvin I understand but it's for a code generator (github.com/bloomberg/crane) so I'm trying to make it fool (me)-proof
English
@joomy This shouldn't happen for most recursive data structures because they're almost always trees so you should only need O(lg(n)) stack frames. If you're running into this you must be using a linked list? In which case it's more idiomatic to have 1 class for the list as a whole.
English
@peter0x44 this is for a code generator so I don't want to assume stack size. github.com/bloomberg/crane
English
@guillaumeclaret if you haven't already seen it, this is a great post that helps explain the spectrum: x.com/miike/status/1…
Mike Dodds@miike
English
A difficulty with formal methods is that it can mean many things, from fuzzing (some people consider it to be) to full interactive theorem provers, and even there there are differences.
This is still a gold mine in terms of security.
Charenton-le-Pont, France 🇫🇷 English
@lastland0 @disconcision @KleeneAlgebra someone please make an alternative Claude Code / Codex UI with EU4 aesthetics, this music, and map/panel clicking sounds
English
YouTube
English
people yearn for the tamagotchi
Zach@zherring
I can't believe there's no simple e-ink chess board in market. I've replaced all endless scroll apps with Chess.com. I love the headspace chess (even poorly) puts me in, but I dislike being attached to my phone to do it. A $100 standalone gadget like this would be so nice to have around.
English