Our Blog

OpenLDK: Bringing the Java World into Common Lisp open_in_new

Regent's Ground-Effect Vehicle Reimagines Coastal Travel open_in_new

LISPy things you can do in 64K bytes of core (2019) open_in_new

Assembly Nights (2021) open_in_new

Postmortem: The Tuesday Our AI Bill Had No Owner open_in_new

I've read a lot of postmortems. The best ones aren't about blame. They're about the gap between what...

Valkey 9.2's new rdbcompression lz4 setting cuts a 1M-row RDB from 178MB to 35MB open_in_new

Valkey 9.2.0-rc1's whole-stream LZ4 RDB compression shrank a million-row JSON dataset from about 178MB to 35MB, an 80% cut the old per-string LZF default barely touches. It does nothing for replication unless a second, undocumented-in-the-release-notes setting is also flipped.

Let It Die: The Lazy Email Agent on TabPFN, Gemma and Temporal open_in_new

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady ...

I Earned a Seat at GitHub Universe. It'll Stay Empty, and That's Not Fair open_in_new

Why Am I Writing This? I'm writing this to show how unfair the world we live in can be,...

👻 The $249 Ghost Sale open_in_new

Before I get into it, I just want to say that my recent unemployment has left me a vast amount of...

MacOS Extensions Pack for the Apple Network Server v1.0 open_in_new

Three Times My Test Failed and the LLM Was Right open_in_new

A test fails, and the reflex is: the code is wrong. Debug the code, fix the code, move on. That...

Virtual and Real World Combine at LAN Party (1998) open_in_new

My Passion for Coding Isn't 404. It's 503. open_in_new

Last month, I signed up for a hackathon. I opened a fresh file and named it ideas.md. Forty-eight...

An examination of lost firefly species of the US and Canada open_in_new

From Ancient Epigraphy to 3D Proteins: Exploring google-deepmind/science-skills & Building Custom Plugins in Google Antigravity open_in_new

Exploring the open-source (Apache 2.0 / CC-BY 4.0) architecture of Google Antigravity's Science Plugin (v1.3.1) and Google Cloud MCP endpoints (Vertex AI, BigQuery, Cloud Run) with a 3D WebGL Biomolecular Studio and 3 custom plugins.

Decompiling Fast Food Tycoon 2: Drawing the World open_in_new

I'm Not on the DEV Team. I Still Debugged Their Missing Badges From Public Code. (And My First Test Was Wrong.) open_in_new

I’m not on the DEV team. I don’t have an admin panel, production database access, or internal Slack....

From confession to conspiracy: How the Wedgewood fake polls story spread open_in_new

👾 Nested Bookmarks Organizer for Chrome w/Custom Ticker open_in_new

✨ Finally, some positive excitement! ✨ I went on a tiny mission last night to build something I...

Everyone Clicks 'I Agree.' No One Reads Them: I Fine-Tuned a 4B Model with Tinker to Catch Unfair Terms of Service in 30 Seconds open_in_new

A 9,000-word contract scanned in 30 seconds. How I fine-tuned Qwen 4B with Tinker on the LexGLUE UNFAIR-ToS benchmark to catch arbitration waivers, unilateral changes, and liability caps.

Gemma 4 on a 4 GB GPU in Plain PyTorch: From Out of Memory to Faster Than llama.cpp open_in_new

On a 4 GB GTX 1650 Ti, Gemma 4 E2B does not fit in 16-bit, and Transformers' own 4-bit path runs out of memory on its first forward pass. Kept in Google's QAT 4-bit end to end, with a few Triton kernels and a CUDA graph, plain PyTorch decodes at 88 tok/s, 1.07x llama.cpp on the same card, and holds the full 131,072-token context.

Why birds can fly over Mount Everest (2020) open_in_new

LimonOS Custom x86-64 Kernel open_in_new

188-Year-Old Mystery of Photography's First Human Subjects Solved open_in_new