Fuck you people. Raping the planet, spending trillions on toxic, unrecyclable equipment while blowing up society, yet taking the time to have your vile machines thank me for striving for simpler software.
Just fuck you. Fuck you all.
I can't remember the last time I was this angry.
I need a progress bar for something with a wildly varying estimated time to completion —
but I still want to show constant progress.
So here’s the 💡: a dot moving at constant speed along a space-filling curve (Hilbert?) and dynamically adjusting the curve’s resolution as the ETC changes.
Thoughts?
A friend gently reminded me on my birthday: “There’s always a number base where you’re in your 20s — and the number of such bases keeps increasing.”
I’m more in my 20s than ever!
It’s a special kind of sadness seeing the entire genre of art made with computers that I have been pursuing since the 70s ruined by AI slop. I have never been less motivated to share new work in my life. 😿
There is no longer the satisfaction that your consumer is a human (who might write a comment, say thanks, post on HN, etc.); nor the cred that comes from people reading and remembering the author, posting on HN, etc.
That is, it totally destroys the incentive structure.
Okay. A lot of « optimizations » are about tricking our brains into thinking they do something good for us.
Are there groups looking into designing around our flawed brains rather than against them?
Almost all chat software automatically scroll to the end of the conversation.
(Even terms and repls.)
But not chat AI UIs. Because you are expected to read the output.
Our chat UIs makes it too easy to skip and ignore.
Our new paper, “Datalog with First-Class Facts,” will appear at VLDB ‘25. We are very excited about this work, which harmonizes concepts from functional programming and databases.
Not only that #Clojure have been supporting high performance computing on Apple M CPU since Neanderthal 0.53.0, but the even more optimized operations through Apple Accelerate is coming soon!
Check it out now at github.com/uncomplicate/neanderthal#MachineLearning#Python#NVIDIA#Swift
🖥️ Jeffy, Lead Engineer at Holiday Pirates, walks us through how he introduced Clojure at scale – starting from zero experience to powering 95% of backend development.
🧰 For CTOs and tech leads, it’s a goldmine of insight: www.freshcodeit.com/podcast/the-...
Me: Hi AI. Add this button for me
AI: Yes, I added the button for you. I also deleted a large chunk of your business logic.
Me: Why?
AI: I don't know why. Deal with it
The most distressing part of the decline of reading is that reading's core purpose is to not only present facts, perspectives, and information, but to expose us to the mechanisms of thought itself.
This bug was weird. All of a sudden code stops compiling. The only change? The reader now adds end source location metadata. If I add only end column or end line it’s ok. But not both. The failing form? (ns foo (:require [bar :refer [x y]])
Why did it fail for two extra metadata fields? Well the ns form is generated and the x and y symbols were already choke full of metadata. The two additional fields caused the compiled jvm class to be too large.
(Please re-skeet for visibility) Did you use HyperCard back in the day? What was your experience with it? What did you make? How did it change the way you think about yourself?
One more week to propose a talk for the ⭐SciNoj Light⭐ #Clojure conference.
scicloj.github.io/docs/communi...
If anybody is considering a talk, now is the time to talk with us, and we will help you brainstorm your draft proposal.
Structurally accelerated merge gives a nice boost (7x) in worst case making CLJD faster than CLJ even on bad cases.
See updated gist gist.github.com/cgrand/6303e...
The first few HYTRADBOI 2025 talks are ready! I'm so excited to be finally able to tell people about them!
Talk #1: A YJIT interview with Maxime Chevalier-Boisvert
I used ChatGPT to generate a series of formally correct but factually inaccurate statements about an assigned reading.
The students had to explain how they were wrong.
This is because I think AI is better at faking expertise than constructing solid arguments.
Pencil and paper are back! For the analog games readings, I had students write a short response on paper at the start of class.
They were given two prompts, and surprisingly, no one complained.
I was shocked by how good their calligraphy is!