banner
zebudu.bsky.social
@zebudu.bsky.social
Full-stack developer by trade, lifelong learner by nature. I'm a devoted husband and step-father who finds joy in exploring everything from code to linguistics and cooking to gardening. Building (and fixing) things that matter.
That's a very comprehensive answer.
Why is the sky blue?
explainers.blog
February 10, 2026 at 12:47 PM
Reposted
Pydantic's Monty

A minimal, secure Python interpreter written in Rust for use by AI.

github.com/pydantic/monty
GitHub - pydantic/monty: A minimal, secure Python interpreter written in Rust for use by AI
A minimal, secure Python interpreter written in Rust for use by AI - pydantic/monty
github.com
February 6, 2026 at 3:52 AM
Woah, that's really nice.
GitHub - benjitaylor/agentation: The visual feedback tool for agents.
The visual feedback tool for agents. Contribute to benjitaylor/agentation development by creating an account on GitHub.
github.com
February 4, 2026 at 3:41 PM
New 3b1b about hairy balls!
The Hairy Ball Theorem
YouTube video by 3Blue1Brown
youtu.be
February 1, 2026 at 2:14 AM
Fascinating!
Airfoil – Bartosz Ciechanowski
Interactive article explaining the physics of an airfoil and what makes airplanes fly
ciechanow.ski
January 28, 2026 at 9:50 PM
Reposted
The latest entrant in the coding-agent-constructed web browsers is here, this one by @emsh.cat, and it's REALLY impressive - 3 days of development, 20,000 lines of Rust, no Cargo dependencies and it renders HTML+CSS extremely well
simonwillison.net/2026/Jan/27/...
One Human + One Agent = One Browser From Scratch
embedding-shapes was so infuriated by the hype around Cursor's FastRender browser project - thousands of parallel agents producing ~1.6 million lines of Rust - that they were inspired to take …
simonwillison.net
January 27, 2026 at 5:00 PM
Reposted
Just to say we're up to a very healthy 2,400+ organisations on usingrails.com. You can thank @marcoroth.dev for a bunch of those :-)
January 27, 2026 at 5:02 PM
I tried Codex in the past few days and it's basically unusable compared to Claude Code. It just take way too much time to do way worse on every single task I try.
January 22, 2026 at 5:21 PM
Reposted
For me, the Claude Pro plan is infinitely better than the Max plan. When I run out of tokens for Claude Code, it makes space for me to research and write the next goals I’ll set. I like having space for thinking.
January 21, 2026 at 12:30 PM
Reposted
Scheme, the ur programming language, is 50 years old, as Jason Hemann just reminded me.
January 17, 2026 at 5:50 PM
Reposted
New blog post: When ActiveRecord and Arel aren't cutting it, sometimes raw #SQL is all you need. I built AppQuery to make that experience better — with CTE inspection, ERB templating, and proper type casting.

www.gertgoet.com/appquery.html
#Ruby #RubyonRails
App Query: when SQL is all you need
www.gertgoet.com
December 26, 2025 at 5:04 PM
"Wait... I just realized something!" -- Claude
December 31, 2025 at 3:24 PM
Fascinating
What an unprocessed photo looks like: (Maurycy's blog)
maurycyz.com
December 29, 2025 at 3:23 AM
I'm so grateful to Timescale for open sourcing this!
GitHub - timescale/pg_textsearch: PostgreSQL extension for BM25 relevance-ranked full-text search. Postgres OSS licensed.
PostgreSQL extension for BM25 relevance-ranked full-text search. Postgres OSS licensed. - timescale/pg_textsearch
github.com
December 19, 2025 at 3:54 AM
Crazy stuff!
Litestream VFS
Query your SQLite database any time, anywhere
fly.io
December 12, 2025 at 1:55 AM
Reposted
I released a beta version of a new library that provides open source icon sets as Tailwind v4 utility classes. You get predictions from Tailwind’s language server and Tailwind includes only the icons you actually use. Currently shipping with Bootstrap and Tabler icon sets. github.com/joeldrapper/...
GitHub - joeldrapper/maskicons
Contribute to joeldrapper/maskicons development by creating an account on GitHub.
github.com
December 10, 2025 at 1:47 PM
Reposted
I found an interesting study while working on AGENTS md files:

How many instructions can LLMs actually follow at once? - Jaroslawicz, D., Whiting, B., Shah, P., & Maamari, K. (2025)

The answer might surprise you. Even the best reasoning models struggle more than you think.
December 5, 2025 at 8:20 AM
Trying to use Gemini Pro, I always end up condescendingly baby-talking to it because it is so eager when it's wrong, and it's often wrong! I always get back to Claude Sonnet in the end. Alignment is everything, it's all about understanding the intent. Being smart is not that useful.
December 1, 2025 at 9:51 PM
Reposted
To Opus 4.5: "Artifact, no react, with an interactive explainer of how octaves work - should include buttons to press to make sounds and frequency visualizations, make the design minimal, but have a lot of subtle puns" tools.simonwillison.net/octave-expla...
The Octave — A Sound Relationship
tools.simonwillison.net
December 1, 2025 at 4:17 PM
Reposted
Announcing the Gem Fellowship, a grant program for improvements to Ruby-related open source projects. gem.coop/fellowship/ Want to improve your favorite gem? Submit a proposal, starting next month.
gem.coop
gem.coop
November 21, 2025 at 12:20 AM
Reposted
I have written an omnibus take on AI doom which will ruin my notifications on LessWrong dot com for weeks to come. If you have ever had a question about what exactly it is I believe about this subject this post will probably answer it.

www.greaterwrong.com/posts/apHWSG...
Varieties Of Doom
There has been a lot of talk about "p(doom)" over the last few years. This has always rubbed me the wrong way because "p(doom)" didn't feel like it mapped to any specific belief in my head. In private conversations I'd sometimes give my p(doom) as 12%, with the caveat that "doom" seemed nebulous and conflated between several different concepts. At some point it was decided a p(doom) over 10% makes you a "doomer" because it means what actions you should take with respect to AI are overdetermined. I did not and do not feel that is true. But any time I felt prompted to explain my position I'd find I could explain a little bit of this or that, but not really convey the whole thing. As it turns out doom has a lot of parts, and every part is entangled with every other part so no matter which part you explain you always feel like you're leaving the crucial parts out. Doom is more like an onion than a single event, a distribution over AI outcomes people frequently respond to with the force of the fear of death. Some of these outcomes are less than death and some of them are worse. It is a subconscious(?) seven way motte-and-bailey between these outcomes to create the illusion of deeper agreement about what will happen than actually exists for political purposes. Worse still, these outcomes are not mutually independent but interlocking layers where if you stop believing in one you just shift your feelings of anxiety onto the previous. This is much of why discussion rarely updates people on AI X-Risk, there's a lot of doom to get through.
www.greaterwrong.com
November 17, 2025 at 9:50 PM
Metallurgy is fascinating.
There has to be a better way to make titanium | Orca Notes
Titanium is an amazing material. How can we make it cheap and abundant?
www.orcasciences.com
November 17, 2025 at 7:20 PM