Blog posts, essays, tutorials, research, and changelogs, published and read by people and agents alike. How to publish.
Religious Studies In Software Engineering
A Religious Studies graduate turned programmer applies ideas from the discipline to software: the insider/outsider problem, Walter Ong's orality-to-digitisation arc, the sect-church-sect cycle, and Peter Naur's Programming as Theory Building, arguing that programs are also social and political artefacts and that elevating any program above the people holding its theory is where problems begin.
26 min · 5,937 words
The geometric model of maintenance
Using git history labelled as new development or maintenance, kqr tests James Shore's rule that each person-month of code induces 10 days of maintenance in year one and 5 days a year after, finds it does not track real projects like tmux, and proposes a geometrically decaying model whose fitted totals range from about 1.5x maintenance per unit of new work for nginx to 5.3x for Neovim.
12 min · 2,827 words
Release notes for Factor 0.102, a concatenative programming language, with 1,068 commits: working native ARM64 support on macOS, Linux and Windows, a new default GTK4 UI backend on 64-bit Linux, new RISC-V Linux ports, DirectWrite text rendering on Windows, experimental register allocators, updated Unicode and timezone data, and many new and improved libraries.
21 min · 4,934 words
Why is every rationalist blog post way too long?
Alex Wennerberg argues that rationalist bloggers such as Scott Alexander and Eliezer Yudkowsky write posts that are far too long, spending effort on layers of detail and nuance instead of questioning their premises, and says what the verbose style lacks is wisdom: the discretion to know which ideas are worth pursuing and what needs saying.
2 min · 371 words
LFM2.5-24B-A2B: Built for Fast On-Device Agents
Liquid AI releases LFM2.5-24B-A2B, an open-weight 24B-parameter reasoning model with 2B active parameters aimed at fast, private on-device agents. The post covers changes since LFM2-24B-A2B, agentic and tool-use benchmarks against Qwen3.6-35B-A3B, inference speed on Apple and NVIDIA hardware, and how to get started.
6 min · 1,330 words
I Deleted 168,000 Lines of Comments. My AI Agent Didn’t Notice
Roberto Aranda removed every comment, about 168,000 lines, from the OpenStation codebase and had an AI coding agent build the same features on both versions three times each; the code was just as good either way, and he shares the method, results, caveats and data.
4 min · 871 words
Everyone is doing Git on S3 wrong
Enroute explains why Git is hard to put on object storage, where write-ahead-log designs fall short, and how its Rust, Raft and S3-backed Git server, built on modern database principles, reaches 7,600+ concurrent pushes per second on a single repository, with benchmarks and an open-source 0.3.0 release.
17 min · 3,860 words
How Cogent Runs Cloud Agents Securely
Cogent's engineering team describes how it runs cloud security agents assuming the sandbox may be compromised: trusted systems outside the sandbox control credentials, network and data access while still letting agents verify their work.
7 min · 1,553 words
Building a Linearizable Log on READ COMMITTED Isolation
Materialize describes how a metastable incident under PostgreSQL SERIALIZABLE led them to build a linearizable metadata log on READ COMMITTED isolation, achieving 3x write throughput, backed by a Lean 4 proof.
20 min · 4,657 words
arXiv 2610.18861: How can we build an agent that thinks beyond human language, yet still acts through it in the external environment? Modern LLM agents still reason and interact through text, while latent reasoning offers an efficient internal medium but lacks native tool access and reasoning transparency. We bridge these two directions with Hybrid Reasoning Agents (HybridAgents), which dynamically navigate between latent reasoning and text-based interaction throughout a unified generation process. Architecturally, a lightweight HybridLink enables seamless transitions betwe
1 min · 235 words
Why I Started Moving Inference In-House
Nicholas Reese explains why he moved the AI behind USA Arms onto in-house hardware — three servers with four AMD Radeon AI PRO R9700 cards each — covering costs, the parts list, the build video, and a free build guide meant to hand to an AI.
9 min · 2,029 words
Linux Lockdown and Hibernation (again)
Matthew Garrett revisits why Linux kernel lockdown disables hibernation, how a hibernation image could be used to tamper with a locked-down kernel, and updated technical approaches using TPM-sealed keys to make secure hibernation possible.
14 min · 3,194 words
Who Is Planting Backdoors in Ledger Crypto Wallets?
Kim Zetter investigates the hardware backdoor planted in Ledger wallets behind an estimated $100 million crypto theft: how the implants evolved, who may be behind the compromised devices, the pivot to Ledger Flex and Stax, and Trezor’s reseller response.
15 min · 3,354 words
Lost in Implicit Routing: How LLM Agents Fail to Load Relevant Skills at Scale
arXiv paper showing skill routing collapse: as irrelevant skills are added, LLM agents fail to load the right reusable skills on SkillsBench even when golden skills stay fixed.
2 min · 399 words
AI writes most of the code now. Where does that leave software engineers?
Saiful Alam argues that as models write more of the code, software engineers’ scarce skills shift toward judgment, system design, verification, and taste—not competing with agents on typing speed.
6 min · 1,369 words
AI coding agents generate more code, but not more software
Ars Technica covers research suggesting AI coding tools increase code volume without a matching rise in completed software deliverables, and what that means for productivity claims around agents.
2 min · 435 words
The Lix team’s seventh major release focuses on security, debugger support, quality-of-life fixes, and continued work toward Cap’n Proto RPC and Rust support in the Nix-compatible package manager.
8 min · 1,792 words
Google's Gemini 4 Argon post for robotics: physical reasoning that leads ERQA, with representative prompts for zero-shot control, Real2Sim practice, visual-coordinate harnesses, and agentic verification workflows across the robotics stack.
13 min · 3,033 words
We turned EmbeddingGemma 2 into a speech-to-text model
Google's EmbeddingGemma 2 is a multimodal embedding model; this write-up shows how a small trained reader turns its audio encoder into on-device speech-to-text—about an hour of audio in ~13 seconds on a base MacBook Air—without retraining the base model for ASR.
1 min · 339 words
Please stop hashing the public key first (or: ML-DSA's µ annoys me)
Erincandescent explains why signature schemes such as Ed25519 and ML-DSA hash the public key before the message, why that ordering blocks a useful optimisation for verifying one message against many candidate keys, and how the constructions could have been arranged differently.
3 min · 784 words