1. Steam Machine (store.steampowered.com)
Your games on the big screen
2. Deno Desktop (docs.deno.com)
Build self-contained desktop applications from a Deno project, with framework auto-detection, hot reload, native windowing, auto-update, and cross-platform distribution.
3. Pledging Another $400k to the Zig Software Foundation (mitchellh.com)
4. Never Give Them Your Face (nevergivethemyourface.com)
They are not verifying your age. They are building a permanent identity checkpoint over public life. Here is why to refuse.
6. Unsloth GLM-5.2 – How to Run Locally (unsloth.ai)
Run the new GLM-5.2 model by Z.ai on local hardware!
8. Help I Accidentally a Wigglegram (lmao.center)
9. GLM 5.2 vs. Opus (techstackups.com)
10. Codex logging bug may write TBs to local SSDs (github.com)
Codex SQLite feedback logs can write ~640 TB/year and rapidly consume SSD endurance Issue Codex is continuously writing a large amount of data to the local SQLite feedback log database: ~/.codex/lo...
11. Danish privacy activist Lars Andersen raided by police (xcancel.com)
Police go directly for the circuit breaker panel to avoid being filmed
This post will be in English, because there apparently is a lot of interest in what happened to me yesterday.
I
This post will be in English, because there apparently is a lot of interest in what happened to me yesterday.
I
12. There is minimal downside to switching to open models (www.marble.onl)
13. The Deadly Rise of Giant Trucks and S.U.V.s (www.nytimes.com)
15. Claude Code's "extended thinking" is a summary- not authentic thinking (patrickmccanna.net)
16. Jobs and Software Is Fucked (urflow.bearblog.dev)
Or how I decided to scream into the void about how awful the job market is. I'm still in a very sour mood after the latest rounds of interviews but ultimatel...
17. Moebius: 0.2B image inpainting model with 10B-level performance (hustvl.github.io)
Moebius: 0.2B Lightweight Image Inpainting Framework with 10B-Level Performance
18. Nearly Half of LG Smart TV Apps Contain Residential Proxy SDKs (spur.us)
Spur scanned 6,038 LG and Samsung smart TV apps and found 2,058 with residential proxy SDKs, exposing privacy and home network risks.
19. Show HN: Got sick of ads, so I made my own logic puzzle site (puzzlelair.com)
Puzzle Lair brings Sudoku, Killer Sudoku, Calcudoku, X-Sudoku, Kakuro, Ripple Effect, Crossmath and Nonogram together in one clean, free site. Solve daily logic puzzles, track your stats and climb the leaderboards.
20. Alan Greenspan Dies at 100; Led Fed During Boom Before 2008 Bust (www.bloomberg.com)
22. Why Drawing Tablet Brands Won't Collaborate on Linux Floss Drivers (www.davidrevoy.com)
23. Git is forever. I'm building Oak anyways (oak.space)
Notes from building Oak — version control for agents. Branch-per-session workflows, team and project scoping, and lazy mounts for agent-driven development.
24. Optocam Zero: a Pi Zero based digital camera made using off the shelf components (github.com)
Optocam Zero is a Raspberry Pi Zero based compact digital camera made using off the shelf components. - dorukkumkumoglu/optocamzero
25. Mexico Just Showed Off a New Cheap, Government-Backed EV (gizmodo.com)
Whether it'll be allowed across the border remains to be seen.
26. Good results fine tuning a local LLM like Qwen 3:0.6B to categorize questions (www.teachmecoolstuff.com)
<p>As a fun personal project, I have been working on a chatbot for answering general questions about my household on anything from maintenance questions to doctor’s appointments. </p>
<p>
The general idea is that the chatbot will get its household knowledge through RAG from querying a vector database, but for better results I have made the vector searches metadata aware.
</p>
<p>
Basically, I am running questions through a pre-processing step to categorize questions into known metadata categories (e.g. pool, car, hvac, cooking). The main goal of this is to narrow down the search space for vector ranking to only indexed entries that match the category of the question. As an example, the question “When did we replace our pool pump?” will be mapped to a category called “pool” before querying the Index database.
</p>
<p>
The general idea is that the chatbot will get its household knowledge through RAG from querying a vector database, but for better results I have made the vector searches metadata aware.
</p>
<p>
Basically, I am running questions through a pre-processing step to categorize questions into known metadata categories (e.g. pool, car, hvac, cooking). The main goal of this is to narrow down the search space for vector ranking to only indexed entries that match the category of the question. As an example, the question “When did we replace our pool pump?” will be mapped to a category called “pool” before querying the Index database.
</p>
27. Munich 1991: The Roots of the Current AI Boom (people.idsia.ch)
28. Show HN: Monolisa v3 – a typeface for developers and creatives (www.monolisa.dev)
A font family designed for software developers. Font follows function.
29. Minecraft: Java Edition 26.2, the first version with Vulkan 1.2 (www.minecraft.net)
Minecraft Java Edition 26.2
